Principles Of Soft Computing Using Python

Programming Learn How To Deploy Soft Computing

Models In Real World Applications Gypsy Nandi
download
https://ebookbell.com/product/principles-of-soft-computing-using-
python-programming-learn-how-to-deploy-soft-computing-models-in-
real-world-applications-gypsy-nandi-53990510

Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.

Principles Of Soft Computing Using Python Programming Nandi Gypsy

https://ebookbell.com/product/principles-of-soft-computing-using-
python-programming-nandi-gypsy-54007112

Principles Of Soft Computing 2nd Ed With Cd 2nd S N Sivanandam

https://ebookbell.com/product/principles-of-soft-computing-2nd-ed-
with-cd-2nd-s-n-sivanandam-22047644

Principles Of Soft Computing 3rd Edition S N Sivanandam S N Deepa

https://ebookbell.com/product/principles-of-soft-computing-3rd-
edition-s-n-sivanandam-s-n-deepa-43851648

Principles Of Softmatter Dynamics Basic Theories Noninvasive Methods

Mesoscopic Aspects 1st Edition Rainer Kimmich Auth

https://ebookbell.com/product/principles-of-softmatter-dynamics-basic-
theories-noninvasive-methods-mesoscopic-aspects-1st-edition-rainer-
kimmich-auth-4205708
Biomechanics Of Soft Tissues Principles And Applications First Edition
Al Mayah

https://ebookbell.com/product/biomechanics-of-soft-tissues-principles-
and-applications-first-edition-al-mayah-6992720

Bottomup Selforganization In Supramolecular Soft Matter Principles And

Prototypical Examples Of Recent Advances 1st Edition Stefan C Mller

https://ebookbell.com/product/bottomup-selforganization-in-
supramolecular-soft-matter-principles-and-prototypical-examples-of-
recent-advances-1st-edition-stefan-c-mller-5234902

Principles Of Entrepreneurship In The Industry 4 0 Era Rajender Kumar

https://ebookbell.com/product/principles-of-entrepreneurship-in-the-
industry-4-0-era-rajender-kumar-44864416

Principles Of Entrepreneurship In The Industry 40 Era Rajender Kumar

https://ebookbell.com/product/principles-of-entrepreneurship-in-the-
industry-40-era-rajender-kumar-44875364

Principles Of Macroeconomics Version 90 9th Edition John B Taylor

https://ebookbell.com/product/principles-of-macroeconomics-
version-90-9th-edition-john-b-taylor-44887330
Principles of Soft Computing Using Python
Programming
 IEEE Press
445 Hoes Lane
Piscataway, NJ 08854
IEEE Press Editorial Board
Sarah Spurgeon, Editor in Chief

Jón Atli Benediktsson Behzad Razavi Jeffrey Reed

Anjan Bose Jim Lyke Diomidis Spinellis
James Duncan Hai Li Adam Drobot
Amin Moeness Brian Johnson Tom Robertazzi
Desineni Subbaram Naidu Ahmet Murat Tekalp
Principles of Soft Computing Using Python
Programming

Learn How to Deploy Soft Computing Models in

Real World Applications

Gypsy Nandi
Assam Don Bosco University
Guwahati, India
Copyright © 2024 by The Institute of Electrical and Electronics Engineers, Inc.
All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any
form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise,
except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without
either the prior written permission of the Publisher, or authorization through payment of the
appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,
MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to
the Publisher for permission should be addressed to the Permissions Department, John Wiley &
Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at
http://www.wiley.com/go/permission.

Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley &
Sons, Inc. and/or its affiliates in the United States and other countries and may not be used
without written permission. All other trademarks are the property of their respective owners.
John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best
efforts in preparing this book, they make no representations or warranties with respect to the
accuracy or completeness of the contents of this book and specifically disclaim any implied
warranties of merchantability or fitness for a particular purpose. No warranty may be created or
extended by sales representatives or written sales materials. The advice and strategies contained
herein may not be suitable for your situation. You should consult with a professional where
appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other
commercial damages, including but not limited to special, incidental, consequential, or other
damages.

For general information on our other products and services or for technical support, please
contact our Customer Care Department within the United States at (800) 762-2974, outside the
United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in
print may not be available in electronic formats. For more information about Wiley products,
visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data Applied for:

Hardback ISBN: 9781394173136

Cover Design: Wiley

Cover Image: © Tuomas A. Lehtinen/Getty Images

Set in 9.5/12.5pt STIXTwoText by Straive, Chennai, India

 v

Contents

About the Author xi

Preface xiii

1 Fundamentals of Soft Computing 1

1.1 Introduction to Soft Computing 1
1.2 Soft Computing versus Hard Computing 2
1.3 Characteristics of Soft Computing 4
1.4 Components of Soft Computing 7
1.4.1 Fuzzy Computing 7
1.4.2 Neural Network 10
1.4.3 Evolutionary Computing 12
1.4.4 Machine Learning 19
1.4.5 Other Techniques of Soft Computing 29
Exercises 31

2 Fuzzy Computing 35
2.1 Fuzzy Sets 37
2.1.1 Features of Fuzzy Membership Functions 38
2.2 Fuzzy Set Operations 41
2.3 Fuzzy Set Properties 42
2.4 Binary Fuzzy Relation 45
2.5 Fuzzy Membership Functions 46
2.6 Methods of Membership Value Assignments 49
2.7 Fuzzification vs. Defuzzification 58
2.8 Fuzzy c-Means 62
Exercises 71
vi Contents

3 Artificial Neural Network 75

3.1 Fundamentals of Artificial Neural Network (ANN) 76
3.2 Standard Activation Functions in Neural Networks 81
3.2.1 Binary Step Activation Function 81
3.2.2 Linear Activation Function 83
3.2.3 Sigmoid/Logistic Activation Function 84
3.2.4 ReLU Activation Function 85
3.2.5 Tanh Activation Function 87
3.2.6 Leaky ReLU Activation Function 88
3.2.7 SoftMax Activation Function 90
3.3 Basic Learning Rules in ANN 91
3.3.1 Hebbian Learning Rule 93
3.3.2 Perceptron Learning Rule 93
3.3.3 Delta Learning Rule 94
3.3.4 Correlation Learning Rule 94
3.3.5 Competitive Learning Rule 95
3.3.6 Outstar Learning Rule 95
3.4 McCulloch–Pitts ANN Model 96
3.5 Feed-Forward Neural Network 98
3.5.1 Single-Layer Perceptron 99
3.5.2 Multilayer Perceptron 103
3.5.3 Radial Basis Function Network 107
3.6 Feedback Neural Network 111
3.6.1 Self-Organizing Map (SOM) 111
3.6.2 Hopfield Neural Network (HNN) 116
Exercises 119

4 Deep Learning 123

4.1 Introduction to Deep Learning 123
4.2 Classification of Deep Learning Techniques 125
4.2.1 Convolutional Neural Networks 125
4.2.2 Recurrent Neural Network (RNN) 137
4.2.3 Generative Adversarial Network (GAN) 144
4.2.4 Autoencoders 149
Exercises 155

5 Probabilistic Reasoning 159

5.1 Introduction to Probabilistic Reasoning 159
5.1.1 Random Experiment 160
5.1.2 Random Variables 160
 Contents vii

5.1.3 Independence 161

5.1.4 Sample Space 162
5.1.5 Odds and Risks 162
5.1.6 Expected Values 165
5.2 Four Perspectives on Probability 165
5.2.1 The Classical Approach 166
5.2.2 The Empirical Approach 166
5.2.3 The Subjective Approach 167
5.2.4 The Axiomatic Approach 167
5.3 The Principles of Bayesian Inference 168
5.4 Belief Network and Markovian Network 171
5.4.1 Syntax and Semantics 172
5.4.1.1 Belief Network 172
5.4.1.2 Markovian Network 172
5.4.2 Conditional Independence 172
5.4.3 Learning Methods of the Networks 177
5.5 Hidden Markov Model 178
5.6 Markov Decision Processes 186
5.7 Machine Learning and Probabilistic Models 191
Exercises 194

6 Population-Based Algorithms 197

6.1 Introduction to Genetic Algorithms 197
6.2 Five Phases of Genetic Algorithms 198
6.2.1 Population Initialization 198
6.2.2 Fitness Function Calculation (Evaluation) 199
6.2.3 Parent Selection 200
6.2.4 Crossover 204
6.2.5 Mutation 205
6.3 How Genetic Algorithm Works? 207
6.4 Application Areas of Genetic Algorithms 212
6.4.1 Using GA in Travelling Salesman Problem 212
6.4.2 Using GA in Vehicle Routing Problem 216
6.5 Python Code for Implementing a Simple Genetic Algorithm 221
6.6 Introduction to Swarm Intelligence 225
6.7 Few Important Aspects of Swarm Intelligence 227
6.7.1 Collective Sorting 228
6.7.2 Foraging Behavior 228
6.7.3 Stigmergy 229
6.7.4 Division of Labor 229
viii Contents

6.7.5 Collective Transport 230

6.7.6 Self-Organization 231
6.8 Swarm Intelligence Techniques 233
6.8.1 Ant Colony Optimization 233
6.8.1.1 How ACO Technique Works? 233
6.8.1.2 Applying ACO to Optimization Problems 235
6.8.1.3 Using ACO in Travelling Salesman Problem 236
6.8.1.4 Python Code for Implementing ACO in TSP 239
6.8.2 Particle Swarm Optimization 241
6.8.2.1 How PSO Technique Works? 242
6.8.2.2 Applying PSO to Optimization Problems 243
6.8.2.3 Using PSO in Job-Shop Scheduling Problem 246
6.8.2.4 Python Code for Implementing PSO 247
Exercises 251

7 Rough Set Theory 255

7.1 The Pawlak Rough Set Model 255
7.1.1 Basic Terms in Pawlak Rough Set Model 256
7.1.2 Measures of Rough Set Approximations 260
7.2 Using Rough Sets for Information System 262
7.3 Decision Rules and Decision Tables 263
7.3.1 Parameters of Decision Tables 264
7.3.1.1 Consistency Factor 264
7.3.1.2 Support and Strength 265
7.3.1.3 Certainty Factor 265
7.3.1.4 Coverage Factor 266
7.3.2 Probabilistic Properties of Decision Tables 266
7.4 Application Areas of Rough Set Theory 268
7.4.1 Classification 268
7.4.2 Clustering 271
7.4.3 Medical Diagnosis 273
7.4.4 Image Processing 276
7.4.5 Speech Analysis 278
7.5 Using ROSE Tool for RST Operations 280
7.5.1 Attribute Discretization 280
7.5.2 Finding Lower and Upper Approximations 281
Exercises 285

8 Hybrid Systems 289

8.1 Introduction to Hybrid Systems 289
8.2 Neurogenetic Systems 291
 Contents ix

8.2.1 GA-based Weight Determination of Multilayer Feed-forward Net 292

8.2.2 Neuroevolution of Augmenting Topologies (NEAT) 298
8.3 Fuzzy-Neural Systems 302
8.3.1 Fuzzy Neurons 303
8.3.2 Adaptive Neuro-fuzzy Inference System (ANFIS) 307
8.4 Fuzzy-Genetic Systems 311
8.5 Hybrid Systems in Medical Devices 315
Exercises 322

Index 327
 xi

About the Author

Dr. Gypsy Nandi currently holds the position

of Associate Professor and Head of the Depart-
ment of Computer Applications at Assam Don
Bosco University, located in Assam, India.
With a profound educational background that
includes a PhD in computer science, she
has amassed nearly two decades of invalu-
able experience within the academic sphere.
Driven by a fervent passion for cutting-edge
technologies, she has been instrumental in
advancing the fields of machine learning, data
science, and social network analysis through
her significant contributions.
Her accomplishments extend to successfully managing various government-
sanctioned consultancy and research-based projects. Additionally, she has
authored two impactful books that delve into the domains of data science and soft
computing. She has also secured an Indian patent grant for her innovative design
of a versatile, multi-functional robot.
Over the course of her illustrious academic career spanning 18 years, she has
been a sought-after speaker for both national and state-level events, where she
shares her expertise in her respective fields. Her extensive research output is
evident in her numerous publications, which include esteemed journal articles,
conference papers, and book chapters.
Beyond her academic pursuits, she remains actively engaged in social commit-
ment activities. She serves as the coordinator of VanitAgrata, a women empow-
erment cell at Assam Don Bosco University. Through this initiative, she provides
free digital literacy training to girls and women in rural areas, contributing to the
advancement of society. She has also received international recognition from a
xii About the Author

university in the Philippines for her dedication to service-learning at the institu-

tional level. Her commitment extends to offering free digital literacy training to
various underprivileged communities in rural Assam.
In summary, she stands as a distinguished scholar and educator who has left an
indelible mark on the fields of computer science and technology. Her accomplish-
ments showcase not only academic excellence but also a profound dedication to
social development and enriching the lives of students and communities alike.
 xiii

Preface

In an era defined by rapid technological advancements, the field of soft computing

has emerged as a powerful paradigm for solving complex real-world problems.
Soft computing leverages the principles of human-like decision-making, allowing
machines to handle uncertainty, vagueness, and imprecision in data and rea-
soning. This interdisciplinary field encompasses a variety of computational
techniques, each with its unique strengths and applications.
This comprehensive textbook, Principles of Soft Computing Using Python
Programming, is designed to provide students, researchers, and practitioners with
a solid foundation in the core concepts and techniques of soft computing. With a
focus on clarity and accessibility, this book takes you on a journey through the
fundamental principles and methods that underpin soft computing.
Chapter 1 – Fundamentals of Soft Computing initiates our exploration,
setting the stage by introducing soft computing and distinguishing it from its
counterpart, hard computing. It delves into the key characteristics of soft com-
puting and explores its essential components, including fuzzy computing, neural
networks, evolutionary computing, machine learning, and other techniques.
Engaging exercises at the end of the chapter invite you to apply your newfound
knowledge.
Chapter 2 – Fuzzy Computing delves deeper into one of the cornerstone tech-
niques of soft computing. It covers fuzzy sets, operations on fuzzy sets, properties,
and more. You will also explore the practical aspects of fuzzy computing, such as
membership functions, fuzzification, defuzzification, and the application of fuzzy
c-means clustering.
Chapter 3 – Artificial Neural Network introduces the fundamentals of
artificial neural networks (ANNs), a powerful tool inspired by the human brain.
You will learn about standard activation functions, basic learning rules, and var-
ious types of neural network architectures, including feedforward and feedback
networks. Engaging exercises will help reinforce your understanding of ANN
concepts.
xiv Preface

Chapter 4 – Deep Learning delves into the realm of deep neural networks,
which have revolutionized fields such as computer vision, natural language
processing, and speech recognition. This chapter provides an overview of deep
learning techniques, including convolutional neural networks (CNNs), recur-
rent neural networks (RNNs), generative adversarial networks (GANs), and
autoencoders.
Chapter 5 – Probabilistic Reasoning explores the world of probability and
its applications in soft computing. You will delve into random experiments, ran-
dom variables, and different perspectives on probability. Bayesian inference, belief
networks, Markovian models, and their applications in machine learning are also
covered.
Chapter 6 – Population Based Algorithms introduces genetic algorithms and
swarm intelligence techniques. You will discover how genetic algorithms work
and explore their applications in optimization problems. Additionally, you will
dive into swarm intelligence methods, including ant colony optimization (ACO)
and particle swarm optimization (PSO), with practical Python code examples.
Chapter 7 – Rough Set Theory delves into the Pawlak Rough Set Model and its
applications in information systems, decision rules, and decision tables. You will
explore the use of rough sets in various domains such as classification, clustering,
medical diagnosis, image processing, and speech analysis.
Chapter 8 – Hybrid Systems concludes our journey by discussing hybrid sys-
tems that combine different soft computing techniques, including neuro-genetic
systems, fuzzy-neural systems, and fuzzy-genetic systems. You will also explore
their applications in medical devices.

Each chapter in this book is carefully structured to provide a clear understand-

ing of the topic, with practical exercises to reinforce your learning. Whether you
are a student, researcher, or practitioner, “Principles of Soft Computing Using
Python Programming” equips you with the knowledge and skills to tackle com-
plex real-world problems using the power of soft computing. So, let us embark on
this enlightening journey through the world of soft computing.

16 October 2023 Dr. Gypsy Nandi

Guwahati, Assam, India
 1

Fundamentals of Soft Computing

Soft computing is a vital tool used for performing several computing operations.
It uses one or more computational models or techniques to generate optimum
outcomes. To understand this concept, let us first clarify our idea about computa-
tion. In any computation operation, inputs are fed into the computing model for
performing some operations based on which results are accordingly produced.
In the context of computing, the input provided for computation is called
an antecedent, and the output generated is called the consequence. Figure 1.1
illustrates the basics of any computing operation where computing is done using
a control action (series of steps or actions). Here, in this example, the control
action is stated as p = f (q), where “q” is the input, “p” is the output, and “f ” is
the mapping function, which can be any formal method or algorithm to solve a
problem.
Hence, it can be concluded that computing is nothing but a mapping function
that helps in solving a problem to produce an output based on the input provided.
The control action for computing should be precise and definite so as to provide
accurate solution for a given problem.

1.1 Introduction to Soft Computing

Many a time, it has been noticed that no fixed solution can be found for a com-
putationally hard task. In such a case, a precisely stated analytical model may not
work to produce precise results. For this, the soft computing approach can be used
that does not require a fixed mathematical modeling for problem solving. In fact,
the uniqueness and strength of soft computing lie in its superpower of fusing two
or more soft computing computational models/techniques to generate optimum
results.
The concept of soft computing was evolved by Prof. Lofti A. Zadeh (University
of California, USA) in the year 1981. Soft computing, as described by Prof. Zadeh,

Principles of Soft Computing Using Python Programming: Learn How to Deploy Soft Computing Models
in Real World Applications, First Edition. Gypsy Nandi.
© 2024 The Institute of Electrical and Electronics Engineers, Inc. Published 2024 by John Wiley & Sons, Inc.
2 1 Fundamentals of Soft Computing

Figure 1.1 Basic concepts of

computing.

Input Computing: Output

(Antecedent) p = f(q) (Consequence)

is “a collection of methodologies that aim to exploit the tolerance for imprecision and
uncertainty to achieve tractability, robustness, and low solution cost.” Prof. Zadeh
also emphasized that “soft computing is likely to play an increasingly important role
in many application areas, including software engineering. The role model for soft
computing is the human mind.” Soft computing mimics the notable ability of the
human mind to reason and make decisions in an environment of improbability
and imprecision. The principal components of soft computing include fuzzy logic,
neurocomputing, and probabilistic reasoning (PR).
If you are wondering in which areas soft computing is being used in our day-
to-day lives, the simplest and most common examples include kitchen appliances
(rice cookers, microwaves, etc.) and home appliances (washing machines, refrig-
erators, etc.). Soft computing also finds its dominance in gaming (chess, poker,
etc.), as well as in robotics work. Prominent research areas such as data compres-
sion, image/video recognition, speech processing, and handwriting recognition
are some of the popular applications of soft computing.

1.2 Soft Computing versus Hard Computing

If we consider computing from the perspective of computer science, it is consid-

ered a certain task that can be accomplished using computers. Such computing
may require certain software or hardware systems to accomplish the task(s) and
to derive a certain outcome or output. To understand this easily, let us take a
simple example of a self-driving car (named, say, Ziva). Now, the car Ziva is given
the instructions to start moving (say, from point A) and to arrive at a destination
point B. To accomplish this task, two possible cases can be considered, as discussed
below:

Case A: The car Ziva uses a software program to make movement decisions.
The path coordinates for movement decisions are already included in the soft-
ware program with the help of which Ziva can take a predefined path to arrive
at its destination. Now, suppose, while moving, Ziva encounters an obstacle in
the path. In such a case, the software program can direct it to move to either to
 1.2 Soft Computing versus Hard Computing 3

the right, or to the left, or to take a back turn. In this case, the self-driving car
is not modeled to identify the nature and complexity of the obstacle to make a
meaningful and proper decision. In this situation, the computation model used
for the car is deterministic in nature, and the output is also concrete. Undoubtedly,
there is less complexity in solving the problem, but the output is always fixed due
to the rigidness of the computation method.
Case B: The car Ziva uses a software program to make movement decisions.
However, in this case, the complexity of the program is more compared to the
complexity of the program defined in Case A. This is so as the car is much more
involved in complex decision-making. Ziva can now mimic the human brain in
making decisions when any kind of obstacle is met in between its travel.
Ziva, first of all, assesses the type of the obstacle, then decides whether it can
overcome the obstacle by any means, and finally, it keeps track of if any other
alternate path can be chosen instead of overcoming the obstacle found in the
same path. The decision to be taken by Ziva is not very crisp and precise, as there
are many alternative solutions that can be followed to reach destination point B.
For example, if the obstacle is a small stone, Ziva can easily climb up the stone
and continue on the same path, as it will lead to a computationally less-expensive
solution. However, if the obstacle is a big rock, Ziva may choose an alternative to
choose another path to reach the destination point.
Case C: Now, let us consider Case C, in which the software program is written
to let the self-driving car reach its destination by initially listing out all the possible
paths available to reach from source A to destination B. For each path available,
the cost of traveling the path is calculated and accordingly sorted to reach at the
fastest time possible. Finally, the optimum path is chosen, considering the mini-
mum cost as well as considering avoidance of any major obstacle. It can be realized
that Case C appends both the cases of Case A and Case B to inherit approaches
from both cases. It also adds some functionalities to tackle complex scenarios by
choosing an optimum decision to finally reach destination point B.
The above three cases can be summarized (as listed in Figure 1.2) to check the
points of differences among each of these cases. From each of the above three
cases, it can be observed that the nature of computation in each of the three cases
is not similar.
Notice that emphasis is given on reaching the destination point in the first case.
As the result is precise and fixed, the computation of the Case A type is termed
hard computing. Now, in the second case, the interest is to arrive at an approximate
result, as a precise result is not guaranteed by this approach. The computation of
the Case B type is termed soft computing. The third case inherits the properties
of both Case A and Case B, and this part of computing is referred to as hybrid
computing. Thus, computing in perspective of computer science can be broadly
categorized, as shown in Figure 1.3.
4 1 Fundamentals of Soft Computing

Case A Case B Case C

• Guarantee of • May use or may not • All the features
achieving precise use any mathematical mentioned in Case A
result model for problem and Case B are
• No Ambiguity in the solving followed in Case C
control flow • Most of the time • Some additional
• Mathematical approximate solution functionalities are
formulation of the is obtained rather added to tackle
control flow can be than precise solution complex scenarios
derived • Algorithm behavior
may be evolutionary
in nature inspired
from biological
species

Figure 1.2 Summarization of three varying cases of a self-driving car.

Figure 1.3 Classification of computing

Computing (in computer science).
Hard Soft Hybrid
computing computing computing

The choice on which classification of computing should be used relies mainly

on the nature of the problem to be solved. However, it is important that before
choosing any of the computing techniques for problem solving, we should be clear
about the main differences between hard computing and soft computing. Table 1.1
lists a few notable differences between hard computing and soft computing to deal
with real-world problems.
The points of differences listed in Table 1.1 clear out the fact that soft computing
methods are more suitable for solving real-world problems in which ideal mod-
els are not available. To name a few applications that may be solved using soft
computing techniques include signal processing, robotics control, pattern recog-
nition, business forecasting, speech processing, and many more. Recent research
has given a lot of importance to the field of computational intelligence (CI). While
traditional artificial intelligence (AI) follows the principle of hard computing, CI
follows the principle of soft computing.

1.3 Characteristics of Soft Computing

As we understood that soft computing can deal with imprecision, partial truth,
and uncertainty, its applications are varied, ranging from day-to-day applications
to various applications related to science and engineering. Some of the dominant
 1.3 Characteristics of Soft Computing 5

Table 1.1 Important points of differences between soft computing and hard computing.

Sl. no. Hard computing Soft computing

1 Requires a precisely stated analytical Can deal with imprecise models

model
2 Often requires a lot of computation Can solve a problem in
time to solve a problem reasonably less time
3 These techniques commonly use It mostly imitates the model from
arithmetic, science, and computing nature
4 Cannot be used in real-world Suitable for real-world problems
problems for which an ideal model for which an ideal model is not
is not present present
5 Requires full truth to produce Can work with partial truth to
optimum result produce optimum result
6 Needs a precise and accurate Can work in an environment of
environment improbability and imprecision
7 The programs that are written using The soft computing techniques
these techniques are deterministic are developed mainly to get
better results for any
nondeterministic polynomial
(NP)-complete problems
8 Usually, high cost is involved in Low cost is involved in
developing solutions developing solutions

characteristics of soft computing are listed in Figure 1.4, and a brief discussion on
each of these characteristics is given next:

(a) Human expertise: Soft computing utilizes human expertise by framing fuzzy
if–then rules as well as conventional knowledge representation for solving
real-world problems that may consist of some degree of truth or false. In short,
where a concrete decision fails to represent a solution, soft computing tech-
niques work best to provide human-like conditional solutions.
(b) Biologically inspired computational models: Computational learning
models that follow the neural model of the human brain have been studied
and framed for complex problem solving with approximation solutions. A few
such popular neural network models include the artificial neural network
(ANN)-, convolutional neural network (CNN)-, and the recurrent neural
network (RNN)-based models. These models are commonly used for solving
classification problems, pattern recognition, and sentiment analysis.
(c) Optimization techniques: Complex optimization problems that are
inspired by nature are often used as soft computing techniques. For example,
6 1 Fundamentals of Soft Computing

Human
expertise
1.0 Biologically
a) Applicable inspired
to real- computational
world models
problems 2.0
b) 7.0
Characteristics
of soft
computing Optimization
Model-free
techniques
learning
3.0
6.0

Goal-driven Fault
tolerant
5.0
4.0

Figure 1.4 Soft computing characteristics.

genetic algorithms (GA) can be used to select top-N fit people out of a human
population of a hundred people. The selection of the most fit people is done
by using the mutation properties inspired by biological evolution of genes.
(d) Fault tolerant: Fault tolerance of a computational model indicates the
capacity of the model to continue operating without interruption, even if
any software or hardware failure occurs. That is, the normal computational
process is not affected even if any of the software or hardware components fail.
(e) Goal-driven: Soft computing techniques are considered to be goal-driven.
This indicates that emphasis is given more on reaching the goal or destina-
tion than on the path considered to be taken from the current state to reach
the goal. Simulated annealing and GA are good examples of goal-driven soft
computing techniques.
(f) Model-free learning: The training models used in soft computing need not
be already aware of all the states in the environment. Learning of a step takes
place in due course of actions taken in the present state. In other words, it can
be said that there is a teacher who specifies beforehand all the precise actions
to be taken per condition or state. The learning algorithm indirectly only has a
critic that provides feedback as to whether the action taken can be rewarded or
punished. The rewards or punishments given help in better decision-making
for future actions.
 1.4 Components of Soft Computing 7

(g) Applicable to real-world problems: Most of the real-world problems are

built on uncertainties. In such circumstances, soft computing techniques
are often used to construct satisfactory solutions to deal with such real-world
problems.

1.4 Components of Soft Computing

The three principal components of soft computing include fuzzy logic-based com-
puting, neurocomputing, and GA. These three components form the core of soft
computing. There are a few other components of soft computing often used for
problem solving, such as machine learning (ML), PR, evolutionary reasoning, and
chaos theory. A brief summary of all these components of soft computing tech-
niques is explained next, along with an illustrative diagram, as given in Figure 1.5.
While fuzzy computing involves understanding fuzzy logic and fuzzy sets,
neural networks include the study of several neural network systems such as
artificial neural network (ANN) and CNN. Evolutionary computing (EC) involves
a wide range of techniques such as GA and swarm intelligence. Techniques for
ML are categorized mainly as supervised learning (SL), unsupervised learning,
and reinforcement learning (RL). Soft computing also involves a wide variety of
techniques such as chaos theory, PR, and evolutionary reasoning.

1.4.1 Fuzzy Computing

The idea of fuzzy logic was first familiarized by Dr. Lotfi Zadeh of the University
of California at Berkeley in the 1960s. While Boolean logic allows evaluation of
output to either 0 (false) or 1 (true), and no other acceptance values in between,
fuzzy logic, on the other hand, is an approach to computing that works on the basis
of the “degrees of truth” that can consider any values between 0 and 1. That is, fuzzy
logic considers 0 and 1 as extreme values of a fact or truth (value “0” represents
absolute false, and value “1” represents absolute true). Any value between 0 and 1
in fuzzy logic indicates the various levels or states of truth.

Components of
soft computing

Fuzzy Neural Evolutionary Machine Other

computing networks computing learning techniques

Figure 1.5 Components of soft computing.

8 1 Fundamentals of Soft Computing

Fully profitable
Is the XYZ courier service profitable?

Is the XYZ courier service profitable?

(Value: 1)

Moderately profitable
Yes
(Value: 0.75)
(Value: 1)
Neither profitable nor
unprofitable (Value: 0.5)
No
Moderately unprofitable
(Value: 0)
(Value: 0.25)

Fully unprofitable
(Value: 0)

Figure 1.6 (a) Boolean (nonfuzzy) and (b) fuzzy logic-based solutions for a problem.

Let us understand this simple concept with the help of an example. For instance,
if we consider the question, “Is the XYZ Courier Service Profitable?” the reply to
this question can be simply stated as either “Yes” or “No.” If only two close-ended
choices are provided for this question, it can be considered as value 1 if the answer
given is “Yes” or 0 if the answer given is “No.” However, what if the profit is not
remarkably well, and only a moderate profit is incurred from the courier service?
If we have a deeper look at the question, there is a possibility that the answer can
be within a range between 0 and 1, as the amount of profitability level may be not
totally 100% profitable or 100% unprofitable. Here, the role of fuzzy logic comes
into play where the values can be considered in percentages (say, neither profit
nor loss, i.e., 0.5). Thus, fuzzy logic tries to deal with real-world situations, which
consider partial truth as a possible solution for a problem.
Figure 1.6(a) illustrates the two outcomes provided for the question “Is the
XYZ Courier Service Profitable?” The solution provided for the question in
this case is Boolean logic based, as only two extreme choices are provided for
responses. On the other hand, Figure 1.6(b) illustrates the various possibilities of
answers that can be provided for the same question “Is the XYZ Courier Service
Profitable?” Here, the concept of fuzzy logic is applied to the given question by
providing a few possibilities of answers such as “fully unprofitable,” “moderately
unprofitable,” “neither profitable nor unprofitable,” “moderately profitable,” and
“fully profitable.” The class membership is determined by the fuzzy membership
function. As seen in Figure 1.6(b), the membership degree (e.g., 0, 0.25, 0.5, 0.75,
and 1) is taken as output value for each response given.
One common example of using fuzzy sets in computer science is in the field
of image processing, specifically in edge detection. Edge detection is the process
of identifying boundaries within an image, which are areas of rapid-intensity
 1.4 Components of Soft Computing 9

changes. Fuzzy logic can be used to make edge detection more robust and
accurate, especially in cases where the edges are not clearly defined. Let us
consider a grayscale image where each pixel’s intensity value represents its
brightness. To detect edges using fuzzy logic, one might define a fuzzy set for
“edgeness” that includes membership functions like “definitely an edge,” “possibly
an edge,” and “not an edge”. In such a case, the membership functions can be
defined as follows:
(a) Definitely an edge: If the intensity difference is high, the pixel is more likely
to be on an edge.
(b) Possibly an edge: If the intensity difference is moderate, the pixel might be
on an edge.
(c) Not an edge: If the intensity difference is low, the pixel is unlikely to be on
an edge.
Using these membership functions, you can assign degrees of membership to
each pixel for each of these fuzzy sets. For instance, a pixel with a high-intensity
difference would have a high degree of membership in the “definitely an edge”
fuzzy set.
A crisp set, as you may know, is a set with fixed and well-defined boundaries.
For instance, if the universal set (U) is a set of all states of India, a crisp set may be
a set of all states of North-East India for the universal set (U). A crisp set (A) can
be represented in two ways, as shown in Equations (1.1) and (1.2)
A = {a1 , a2 , a3 , … , an } (1.1)
A = {x|P(x)} (1.2)
Here, in Equation (1.2), the crisp set “A” consists of a collection of elements
ranging from a1 to an . Equation (1.2) shows the other way of representing a crisp
set “A,” where “A” consists of a collection of values of “x” such that it has got the
property P(x).
Now, a crisp set can also be represented using a characteristic function, as shown
in Equation (1.3):
{
1, if x belongs to A
𝜇A′ (x) = (1.3)
x, if x does not belong to A
A fuzzy set is a more general concept of the crisp set. It is a potential tool to
deal with uncertainty and imprecision. It is usually represented by an ordered pair
where the first element of the ordered pair represents the element belonging to a
set, and the second element represents the degree of membership of the element
to the set. The membership function value may vary from 0 to 1. Mathematically,
′
a fuzzy set A is represented as shown in Equation (1.4).
{
A′ = x, 𝜇A′ (x)|| xX} (1.4)
10 1 Fundamentals of Soft Computing

Here, the membership function value indicates the degree of belongingness and is
denoted by 𝜇A′ (x). Here, in Equation (1.4), “X” indicates the universal set, which
consists of a set of elements “x.” A membership function can either be any stan-
dard function (for example, the Gaussian function) or any user-defined function
in requirement to the problem domain. As this membership function is used to
represent the degree of truth in fuzzy logic, its value on the universe of discourse
“X” is defined as:
𝜇A′ (x) = [0, 1] (1.5)
Here, in Equation (1.5), each value of “X” represents an element that is mapped
to a value between 0 and 1.
The above explanations lead us to the understanding that a fuzzy set does
not have a crisp, clearly defined boundary; rather it contains elements with
only a partial degree of membership. Some of the standard properties of fuzzy
sets include commutative property, associative property, distributive property,
transitivity, and idempotent property. There are few other properties of fuzzy sets
that will be elaborately discussed in Chapter 2.
Also, there are three standard fuzzy set operators used in fuzzy logic – fuzzy
union, fuzzy intersection, and fuzzy complement. In case of complement oper-
ation, while a crisp set determines “Who do not belong to the set?,” a fuzzy set
determines “How many elements do not belong to the set?” Again, in case of union
operation, while a crisp set determines “Which element belongs to either of the set?,”
a fuzzy set determines “How much of the element is in either of the set?” Lastly, in
case of intersection operation, while a crisp set determines “Which element belongs
to both the sets?,” a fuzzy set determines “How much of the element is in both the
sets?” These fuzzy operations will also be elaborately discussed in Chapter 2.
Fuzzy logic systems have proved to be extremely helpful in dealing with
situations that involve decision-making. As some problems cannot be solved by
simply determining whether it is True/Yes or False/No, fuzzy logic is used to
offer flexibility in reasoning in order to deal with uncertainty in such a situation.
The applications of fuzzy logic are varied, ranging from domestic appliances to
automobiles, aviation industries to robotics.

1.4.2 Neural Network

The human brain consists of billions of interconnected neurons. These neurons
are cells that use biochemical reactions to receive data, and accordingly process
and transmit information. A typical neuron consists of four main parts – dendrites
(receptors that receive signals from other neurons), soma (the cell body that sums
up all the incoming signals to create input), axon (the area through which neu-
ron signals travel to other neurons when a neuron is fired), and synapses (point of
 1.4 Components of Soft Computing 11

Soma (cell body)

Dendrites

Synapse

Axon

Figure 1.7 Parts of a neuron.

interconnection of one neuron with other neurons). The different parts of a neuron
are illustrated in Figure 1.7. A neuron gets fired only if certain conditions are met.
The signals received on each synapse may be of excitatory or inhibitory type.
When the excitatory signals exceed the inhibitory signals by certain quantified
threshold value, the neuron gets fired. Accordingly, either positive or negative
weights are assigned to signals – a positive weight is assigned to excitatory signals,
whereas a negative weight is assigned to inhibitory signals. This weight value indi-
cates the amount of impact of a signal on excitation of the neuron. The signals
multiplied by the weight in all the incoming synapse is summed up to get a final
cumulative value. If this value exceeds the threshold, then the neuron is excited.
This biological model has been mathematically formulated to accomplish optimal
solutions to different problems and is technically termed as “Artificial Neural Net-
work (ANN).” ANN has been applied in a large number of applications such as pat-
tern matching, pattern completion, classification, optimization, and time-series
modeling.
A simple example of an ANN is given in Figure 1.8. The nodes in ANN are orga-
nized in a layered structure (input layer, hidden layer, and output layer) in which
each signal is derived from an input and passes via nodes to reach the output.
Each black circular structure in Figure 1.8 represents a single neuron. The sim-
plest artificial neuron can be considered to be the threshold logic unit (TLU). The
TLU operation performs a weighted sum of its inputs and then outputs either a “0”
or “1.” An output of “1” occurs if the sum value exceeds a threshold value and a
“0” otherwise. TLU thus models the basic “integrate-and-fire” mechanism of real
neurons.
The basic building block of every ANN is the artificial neuron. At the entrance
section of an artificial neuron, inputs are assigned weights. For this, every input
value is multiplied by an individual weight (Figure 1.9). In the middle section of
the artificial neuron, a sum function is evaluated to find the sum of all the weighted
12 1 Fundamentals of Soft Computing

Input Hidden Output

layer layer layer

Input 1

Input 2 Output

Input 3

Figure 1.8 A simple example of neural network.

Sum Transfer function

Multiplication

Input 1 x

Input 2 x ∑ F(X) Output

Bias

Figure 1.9 The design of an artificial neuron.

inputs and bias. Next, toward the exit of the artificial neuron, the calculated sum
value is passed through an activation function, also called a transfer function.
ANN provides a simplified model of the network of neurons that occur in the
human or animal brain. ANN was initially found with the sole purpose of solving
problems in the same way that a human or animal brain does. However, more
and more research on ANN has led to the deviation of ANN from biology to solve
several challenging tasks such as speech recognition, medical diagnosis, computer
vision, and social network filtering.

1.4.3 Evolutionary Computing

EC is a distinct subfield of soft computing that has gained wide popularity in the
past decade in various areas of research related to natural evolution. In the case
of natural evolution, an environment consists of a population of individuals that
struggle for survival and strive for reproduction. The fitness of each individual
decides its probability of being able to survive in a given environment. Evolution-
ary algorithms (EA) follow heuristic-based approach to problem solving, as these
algorithms cannot be solved in polynomial time. Many variants of EC have evolved
 1.4 Components of Soft Computing 13

Initialization Selection Termination

Crossover Mutation

Figure 1.10 Basic steps of evolutionary algorithms.

over time, and each variant is suitable for more specific types of problems and
data structures. At times, two or more evolutionary algorithms (EA) are applied
together for problem solving in order to generate better results. This makes EC
very popular in computer science, and a lot of research is explored in this area.
In general, EA mimic the behavior of biological species based on Darwin’s theory
of evolution and natural selection mechanism. The four main steps involved in EA
include – initialization, selection, use of genetic operators (crossover and mutation),
and termination. Each of these chronological steps makes an important contribu-
tion to the process of natural selection and also provides easy ways to modularize
implementations of EA. The four basic steps of EA are illustrated in Figure 1.10,
which begins with the initialization process and ends with the termination process.
The initialization step of EA helps in creating an initial population of solutions.
The initial population is either created randomly or created considering the ideal
condition(s). Once the population is created in the first step, the selection step is
carried out to select the top-N population members. This is done using a fitness
function that can accurately select the right members of the population. The next
step involves use of two genetic operators – crossover and mutation – to create
the next generation of population. Simply stated, these two genetic operators help
in creating new offspring from the given population by introducing new genetic
material into the new generation. Lastly, the EA involve the termination step to end
the process. The termination step occurs in either of the cases – the algorithm has
reached some maximum runtime, or the algorithm has reached some threshold
value based on performance.
Independent research work on EA led to the development of five main streams
of EA, namely, the evolutionary programming (EP), the evolution strategies
(ES), swarm intelligence, the GA, and the differential evolution (DE) (as shown
in Figure 1.11). A brief discussion on each of these subareas of EA is discussed in
the later part of this section.
● Evolutionary programming: The concept of EP was originally conceived by
Lawrence J. Fogel in the early 1960s. It is a stochastic optimization strategy
14 1 Fundamentals of Soft Computing

Evolutionary
programming
Evolutionary algorithms

Evolutionary Ant colony

strategies optimization

Artificial bee
Swarm intelligence colony

Particle swarm
Genetic algorithms
optimization

Differential evolution

Figure 1.11 Main families of evolutionary algorithms.

similar to GA. However, a study is made on the behavioral linkage of parents and
offspring in EP, while genetic operators (such as crossover operators) are applied
to produce better offspring from given parents in GA. EP usually involves four
main steps, as mentioned below. Step 1 involves choosing an initial population
of trial solutions. Step 2 and Step 3 are repeated either until a threshold value
for iteration exceeds or an adequate solution for the given problem is obtained:
⚬ Step 1: An initial population of trial solutions is chosen at random.
⚬ Step 2: Each solution is replicated into a new population. Each of these off-
spring solutions is mutated.
⚬ Step 3: Each offspring solution is assessed by computing its fitness.
⚬ Step 4: Terminate.
The three common variations of EP include the Classical EP (uses Gaussian
mutation for mutating the genome), the Fast EP (uses the Cauchy distribution
for mutating the genome), and the Exponential EP (uses the double exponential
distribution as the mutation operator). A few of the common application areas
of EP include path planning, traffic routing, game learning, cancer detection,
military planning, combinatorial optimization, and hierarchical system design.
● Evolution strategies: Evolution strategies (ES) is yet another optimization

technique that is an instance of an evolutionary algorithm. The concept of ES

was proposed by three students, Bienert, Rechenberg, and Schwefel, of the
Technical University in Berlin in 1964. ES is also inspired by the theory of
evolution. In fact, it is inspired mainly by the species-level process of evolution
(phenotype, hereditary, and variation). The main aim of ES algorithm is
to maximize the fitness of a group of candidate solutions in the context of
an objective function from a domain. ES usually involves six main steps as
mentioned below, out of which Steps 2–5 are repeated until convergence:
 1.4 Components of Soft Computing 15

– Step 1: Randomly choose n population of individuals.

– Step 2: Create n population of parameters 𝜃 1 , 𝜃 2 , …, 𝜃 n by adding Gaussian
noise to the best parameter (𝜃 is the parameter vector).
– Step 3: Evaluate the objective function for all the parameters, and select the
top-N best-performing parameters (elite parameters).
– Step 4: Find the best parameter (best parameter = mean(top-N elite parame-
ters)).
– Step 5: Decay (minimize) the noise by some factor.
– Step 6: Terminate.
Mutation and selection are two main operations performed in evolution strate-
gies. These two operations are applied continuously until a termination criterion
is met. While selection operation is deterministic and based on fitness rankings,
mutation is performed by adding a normally distributed random value to each
vector component. The simplest case of ES can be a population of size two – the
parent (current point) and the result of its mutation. In such a case, if the fitness
of the mutant is either equal or better than the fitness of the parent, the mutant
then becomes the parent for the next iteration. If not, the mutant is discarded.
● Swarm intelligence: Swarm intelligence (SI) algorithms are a special type
of EA. The concept of SI was first introduced by Beni and Wang in “swarm
intelligence in cellular robotic system.” SI algorithms adapt to the concept of
the behavior of swarms. A swarm is a dense group of homogenous agents
that coordinate among themselves and the environment for an interesting,
combined clustered behavior to emerge. The term swarm is mostly used in
biological concepts to explain the coordinated behavior of a group of animals,
fish, or birds. For example, a colony of ants marching together in search of food
is a remarkable example of clustered behavior found in biological environment.
SI algorithms mainly include particle swarm optimization (PSO), ant colony
optimization (ACO), and artificial bee colony (ABC).
(a) Particle swarm optimization: Particle swarm optimization (PSO) is a
nature-inspired population-based stochastic optimization technique devel-
oped by Kennedy and Eberhart in 1995. PSO algorithms mimic the social
behavior of animals, such as fish schooling and bird flocking, in which fish
or birds move collectively to solve a task. The PSO technique works on the
same principle of foraging behavior of biological species. PSO is easy to
implement, as it requires adjustment of only a few parameters. This is the
reason why PSO has been successfully applied in many application areas,
such as the traveling salesman problem, scheduling problem, sequential
ordering problem, and vehicle routing problem.
(b) Ant colony optimization: Ant colony optimization (ACO) is a population-
based metaheuristic mainly used to find an approximate solution to a given
16 1 Fundamentals of Soft Computing

Obstacle

Food
Nest

Figure 1.12 Colony of ants marching toward food source.

challenging optimization problem. ACO uses a set of software agents called

artificial ants, which help find good solutions to a given optimization prob-
lem. Figure 1.12 shows the concept of how a colony of ants marches together
to reach the food source. In between, if any obstacle is met for the first time,
the ants divide among themselves to travel in both directions. However,
the shorter route is noted and followed by the rest of the ant based on the
pheromone (chemical substance produced and released into the environ-
ment) deposit concentration.
The pheromone smell deposited by the ant on the pathway provides an
indication to the other worker ants about the presence of food in a nearby
area. The pheromone trails help create indirect communication between all
the nearby ants, which helps in finding the shortest path between the food
source and the nest.
(c) Artificial bee colony: Artificial bee colony (ABC) is a computing tech-
nique that is based on the intelligent foraging behavior of honey bee
swarms. This concept was first proposed by Dervis Karaboga of Erciyes
University in the year 2005. The ABC model considers three types of bees:
the employed bees, the onlooker bees, and the scout bees. Usually, only one
artificial employed bee is hired for a food source. The artificial employed
bee targets the food source and performs a dance after returning to its hive.
Once the target food source is over, the employed bee shifts its position to a
scout and starts hunting for a new food source. The role of onlookers is to
watch the dance of employed bees and choose food sources based on the
performance of the dances. The onlookers and the scouts are considered
unemployed bees.
ABC accomplishes the task of hunting for food through social cooper-
ation. In the ABC problem, each food source signifies a possible solution
 1.4 Components of Soft Computing 17

to the optimization problem, and the amount of nectar in the food sources
decides the quality or fitness of the given solution. In fact, the quality of
a food source depends on many factors, such as the amount of food source
available, the ease of extracting its nectar, and also its distance from the nest.
Depending on the number of food sources, the same number of employed
bees is chosen to solve a problem. It is the role of employed bees to carry on
the information about the quality of the food source and share this informa-
tion with the other bees.
The unemployed bees also play an active role in the food hunt. One type
of unemployed bee is the scout, which explores the environment near the
nest in search of food. The other type of unemployed bee is the onlooker,
which waits in the nest to get information about the quality of food sources
from the employed bees and establish the better food sources. Communica-
tion among bees related to the quality of food sources takes place through
the famous “waggle dance” of honey bees. This exchange of information
among the three types of bees is the most vital occurrence in the formation
of collective knowledge.
● Genetic algorithms: The concept of genetic algorithms (GA) was proposed by
John Holland in the 1960s. Later, Holland along with his colleagues and stu-
dents developed the concepts of GA at the University of Michigan in the 1960s
and 1970s as well. A genetic algorithm is a metaheuristic that is inspired by
Charles Darwin’s theory of natural evolution. GA are a part of the larger class of
EA that emphasize on selecting the fittest individuals for reproduction in order
to produce offspring. The generated offspring inherit the characteristics of their
parents and is therefore expected to have better fitness if the parents do have
good fitness values. Such offspring, in turn, will have a better chance of sur-
vival. If this process continues to repeat multiple times, at some point in time, a
generation of the fittest individuals will be formed.
There are basically five main phases of GA (as illustrated in Figure 1.13): pop-
ulation initialization, fitness function calculation, parent selection, crossover,
and mutation. Initially, a random population of size “n” consisting of several
individual chromosomes is chosen. Next, the fitness value of each of the indi-
vidual chromosomes is calculated based on a fitness function. The fitness value
plays a vital role in the decision-making of the selection of chromosomes for
crossover.
In the crossover phase, every two individual chromosomes selected are repro-
duced using a standard crossover operator. This results in the generation of two
offspring from each pair of chromosomes. The new offspring generated are then
mutated to produce a better set of individual chromosomes in the newly gener-
ated population. These entire five phases of GA are repeated until a termination
condition is met. Each iteration of the GA is called a generation, and the entire
18 1 Fundamentals of Soft Computing

Population Fitness Result

initialization evaluation Selection (fittest solution)

Crossover
Mutation

Figure 1.13 Steps followed in genetic algorithms.

set of generations is called a run. The final output (result) is the generation of
the fittest individuals that have the greatest chance of survival.
● Differential evolution: Differential evolution (DE) is a common evolutionary
algorithm stimulated by Darwin’s theory of evolution and has been studied
widely to solve diverse areas of optimization applications since its inception
by Storn and Price in the 1990s. The various steps involved in DE include
population initialization, mutation, crossover, selection, and result generation
(illustrated in Figure 1.14). Prior to applying these basic steps, the parameters
of DE need to be defined, such as the population size, the selection method,
the crossover method, and the perturbation rate (the weight applied to random
differential).
There are various popular variants of DE, some of which are mentioned below:
● The standard differential evolution (DE)
● The self-adaptive control parameters differential evolution (JDE)
● The adaptive differential evolution with optional external archive (JADE)
● The composite differential evolution (CODE)
● The self-adaptive differential evolution (SADE)
The applications of DE are varied, including synthesis and optimization of
heat-integrated distillation system, optimization of an alkylation reaction, digital

Population Fitness
initialization evaluation Mutation

Result Crossover
(fittest solution) Selection

Figure 1.14 Steps followed in differential evolution.

 1.4 Components of Soft Computing 19

filter design, optimization of thermal cracker operation, batch fermentation

process, estimation of heat transfer parameters in bed reactor, engineering system
design applied in multiobjective context, and parameter estimation in fed-batch
fermentation process.

1.4.4 Machine Learning

ML is an emerging area in the field of computer science, and its applications are
varied, be in medical diagnosis or speech recognition, stock market predictions
or gaming, and Google translation or recommendation systems. In fact, ML is a
branch of AI that allows systems to learn through experience rather being explic-
itly controlled through programs. Thus, the core approach in ML is to train a
machine for developing the capability of learning without being explicitly pro-
grammed to provide optimum output.
ML initially follows the procedure of feeding the machine with required input
data and then training the machine to a certain level so that it is able to learn and
adapt to the ability of predicting the output based on the training acquired. Let us
try to understand and analyze this approach with the help of a suitable example,
as given in Figure 1.15. Imagine input data of various objects of different shapes
and colors are fed into the computer system. The ML technique can be applied
to cluster the objects based on the colors, shapes, and sizes. Figure 1.15 shows the
output in which four clusters have been formed after training the model based on
input provided.
Let us consider another example of ML, as shown in Figure 1.16. The emails
received by a user are trained by ML algorithm to classify these emails as either
legitimate mail (sent to inbox folder) or spam emails (sent to spam folder). The ML
algorithm initially uses a dataset that has a list of emails classified either as spam

Cluster 1

Cluster 2
Machine learning
algorithm

Cluster 3

Cluster 4

Figure 1.15 Machine learning algorithm used for training data to form clusters.
20 1 Fundamentals of Soft Computing

Machine learning
algorithm

Figure 1.16 Machine learning algorithm used for classifying email as spam or
legitimate.

or not spam. Once the algorithm gets fully trained and shows high accuracy in
prediction, this algorithm is now all ready to be used for such similar predictions
for the future.
The contribution of ML right from solving day-to-day tasks to solving complex
real-life problems is tremendous. In fact, many home appliances, health care
monitoring systems, mobile apps, and internet services heavily rely on using
ML. Also, popular virtual personal assistants such as Alexa, Siri, and Google Now
rely dominantly on the techniques of ML. These popular virtual assistants are a
perfect example of the usage of advanced ML, as it has a number of capabilities
that include voice interaction, playing audio, answering the door, dimming the
lights of a room, and reading the latest headlines. Again, traffic predictions using
GPS navigation services use ML techniques to provide live data to users regarding
traffic congestion while traveling. ML algorithms also help companies develop
chatboxes to solve user queries. It is expected that the contribution of ML in the
near future will continue exceeding, and researchers have to extensively depend
on such ML algorithms to build innovative tools and techniques.
There are mainly four types of ML – SL, unsupervised learning, semi-supervised
learning, and RL. Under the umbrella of SL are classification and regression,
which use a dataset having known inputs and outputs. Unsupervised learning, on
the other hand, uses a dataset to identify hidden patterns. Under the umbrella of
unsupervised learning falls clustering and association analysis. Semi-supervised
learning lies between SL and unsupervised learning and handles the drawbacks
of both these types of ML techniques.
Figure 1.17 illustrates the various types of ML used for different variety of data
and problems. All these types of ML have a special role to perform, which is
explained in brief next. Let us now try to understand these four types of ML in brief.

● Supervised learning (SL): Most of the ML techniques are based on SL, which
works on the basis of supervision. Basically, supervised learning techniques
train machines using “labeled” datasets. Such datasets have input variables
that are used by a mapping function to derive the output variable(s). This can
be mathematically expressed as P = f (Q), where Q is the input data, and P is the
 1.4 Components of Soft Computing 21

Types of
machine learning

Supervised Semi-supervised Unsupervised Reinforcement

learning learning learning learning

Classification Clustering

Regression Association
analysis

Figure 1.17 Types of machine learning.

output. SL thus uses the input variables and the corresponding output having
“labeled” data to train machines by approximating the mapping function.
The training is carried out accurately to an extent such that when a new input
data (q1) is fed, a perfect prediction can be made to display the output variable
(P) for that data (q1).
To clear the idea of SL, let us consider a simple example. Let us assume that a
dataset that consists of images of three varying animals – say, dog, goat, and cat,
which are provided as input to the machine. The output variable, also called
the labeled variable, stores the values as either “dog,” “goat,” or “cat.” Now, a
SL model is trained with this labeled dataset that can differentiate among these
three animals and correctly predict the output. This learning model is illustrated
in Figure 1.18, in which the labeled data are trained, and the model is tested with
a test dataset (unlabeled data) to check the accuracy of the predicted output once
the training is completed. If the accuracy of prediction is very high, the model
can be set to be trained and ready for use for future predictions.
There are two main notable techniques of supervised ML – classification and
regression. Both these techniques have a similar goal of predicting the output
(dependent attribute) based on the series of input data provided. Classification
deals with prediction of discrete class label(s), whereas regression deals with
prediction of continuous quantity. Also, a classification problem is mainly eval-
uated using accuracy as the evaluation metric, whereas a regression problem is
usually evaluated using root mean square error (RSME).
(a) Classification: Classification is a type of supervised ML as it considers
“labeled” data to perform the task of prediction of output. The “labeled”
22 1 Fundamentals of Soft Computing

Prediction

Dog
Cat
Training Goat
Goat model
Dog
Cat
Dog
Goat
Dog Predicted output
Labeled data Test data

Figure 1.18 Supervised learning.

variables are termed as classifiers, which play a major role in training the
algorithm. It approximates a mapping function (f ) from input variables (Q)
to discrete output variables (P). For a given observation, the primary task of
classification is to predict the value of class. The technique of classification
is applied in many significant areas where new observations are needed to
be categorized, such as spam filtering, face detection, credit approval, fraud
detection, optical character recognition, market segmentation, and so on.
For a given observation, the primary task of classification is to predict the
value of class.
As shown in Figure 1.18, the problem given is of classification in which
the output to be obtained belongs to any of the three classes – cat, dog, or
goat. Another example of a problem of classification would be if the output
to be predicted is either Yes or No, such as “Diabetes” or “No Diabetes,” “Pro-
vide Loan” or “Do not Provide Loan,” “Spam Mail” or “Legitimate Mail,”
and so on.
Classification techniques are further classified under two models – the
lazy learners and the eager learners. In the case of lazy learning model, the
training time taken is comparatively less; however, the prediction time
is more compared to the eager learning model. Examples of lazy learners
include the k-Nearest Neighbor (kNN) classification model, and case-based
learning model. Eager learners construct the classification model from the
given training data even before receiving any test data. Examples of eager
learners include decision trees, Naïve Bayes, and ANNs.
(b) Regression: In regression, the main task is to approximate a mapping
function ( f ) from input variables (X) to a continuous output variable (Y ).
The continuous output variable (Y ) should denote a quantity and has to
be a real value, such as an integer or floating-point value. Given a new
set of input values, regression can make a prediction of the corresponding
quantitative output (y), based on the study of previous corresponding (x, y)
 1.4 Components of Soft Computing 23

paired values. Thus, regression analysis is a predictive modeling technique

that analyses the relationship between a dependent variable (the new y
value) based on the independent variable (the given x value). For example,
given a series of values of the number of sales of tea cups per month at
a tea stall, a prediction on the amount of sale of tea can be made for any
near-future months.
There are various types of regression models, which differ mainly on three
factors – the number of independent variables considered, the shape of the
regression line, and the type of dependent variable being used. If only one
independent variable is used, it is a case of simple regression, whereas if
more than one independent variable is used, it is a case of multiple regres-
sion. Moreover, both simple and multiple regressions can be either linear
or nonlinear. Figure 1.19 mentions the various standard regression models
used for SL for continuous variables.
i. Linear regression: The most common type of regression used is the lin-
ear regression in which a linear relationship is maintained between the
dependent and independent variables. Here, the dependent variable is
continuous, whereas the independent variable(s) can be either discrete
or continuous.
ii. Logistic regression: In logistic regression, the dependent or target vari-
able can belong to only one of the possible binary values, such as 1 or 0,
True or False, Positive or Negative, and so on. Its probability value ranges
between 0 and 1.
iii. Polynomial regression: When the relationship between the depen-
dent and the independent variables is nonlinear in nature, polynomial
regression is used.
iv. Ridge regression: If two or more independent variables are used that
have high intercorrelations among them, it is the case of multicollinear-
ity. Ridge regression is a type of regression that can be best used for
dealing with multicollinearity in multiple regression data.
v. Lasso regression: LASSO stands for least absolute shrinkage and
selection operator. Lasso regression, also called penalized regression,
uses the shrinkage technique to determine the coefficients. Shrinkage

Standard regression
techniques

Linear Logistic Polynomial Ridge Lasso Support vector

regression regression regression regression regression regression

Figure 1.19 The two main types of regression.

24 1 Fundamentals of Soft Computing

is a concept in which the data values are shrunk toward a central or

middle point.
vi. Support vector regression (SVR): This method supports both linear
and nonlinear regressions. It uses the concepts of hyperplanes, kernels,
boundary lines, and support vectors that are also used in the case of sup-
port vector machines (SVM).
● Unsupervised learning: Many of the real-time datasets are not provided
with any labels. Such datasets are not classified beforehand and hence have
no class labels. SL fails to handle such datasets to solve a ML-based task.
This is where unsupervised learning is used to find commonalities among its
input data without the help of any labels or classifications. The primary goal
of unsupervised learning is to discover hidden patterns from the input data
provided.
To understand unsupervised learning, let us consider the same example
given in Figure 1.18. In this Figure, all the data in the dataset are labeled “cat,”
“dog,” and “goat.” Now, imagine that the dataset is not provided with any label.
The primary task of unsupervised learning is to group similar data based on
some features of input data, such as color, size, pattern, and so on. Figure 1.20
illustrates a case of unsupervised learning in which a dataset without having
any labels is fed into the training model to generate similar clusters as output.
As can be seen in Figure 1.20, the dataset consists of several different images
of cats, dogs, and goats. The results obtained are the three clusters that are
grouped based on the common features (similarities) among each animal.

Cluster 1

Train
Generate
model
clusters
Cluster 2

Dataset
(without labels) Cluster 3

Predicted output

Figure 1.20 Unsupervised learning.

 1.4 Components of Soft Computing 25

There are two main notable techniques of unsupervised ML – clustering

and association rule mining. As these are unsupervised learning techniques,
the models are not supervised using any training dataset. Let us now discuss the
concepts of clustering and association rule mining in brief:
(a) Clustering: Clustering is a simple method in which items or objects are
grouped into clusters based on similarities. The similarities are generated
based on the presence or absence of some features among the objects.
A real-time example where clustering is used is in recommender systems of
e-commerce sites, which groups items of similar tastes for a user based on
his/her previous purchasing experiences.
There are various clustering algorithms, and a choice of which algorithm
is to be chosen relies on whether the algorithm can scale to the given input
dataset. The four types of clustering are explained next:
i. Centroid-based clustering: This clustering approach works by repeat-
edly finding the centroid of each cluster until the results become stable.
A classic way to apply centroid-based clustering is the K-means cluster-
ing, where “k” refers to the number of clusters to be formed. Initially,
“k” number of points are chosen at random for selection of centroids.
The rest of the data points are measured and grouped to the nearest cen-
troid based on a distance measure. The centroids are then again deter-
mined based on the cluster formed. The steps are repeated iteratively
until convergence.
ii. Density-based clustering: This clustering approach connects areas of
high density into clusters. While doing so, the outliers are often left out
and may not be included in any of the clusters formed. The distribu-
tions can be of arbitrary shape or size, as the primary approach is to
consider all objects in the same cluster if they belong to dense areas or
connectivities.
iii. Distribution-based clustering: This clustering approach uses the
concepts of probability distribution such as Gaussian distribution or
binomial distribution to form clusters. The basic idea is that as the
distance from the center of the distribution increases, the probability of
a point belonging to that distribution decreases.
iv. Hierarchical clustering: This clustering approach creates a hierarchy
of clusters in the form of a tree. The tree-like structure that is formed to
create several clusters is called dendrogram. The hierarchical clustering
may follow either of the two approaches: agglomerative or divisive.
The agglomerative approach is a bottom-up approach that initially
begins by considering each data item as a single cluster and then
26 1 Fundamentals of Soft Computing

proceeds by merging the closest pair of clusters. The divisive approach,

on the other hand, works just the opposite of the agglomerative
approach. The divisive approach initially begins by considering only one
big cluster that consists of all the data points. It gradually then splits the
big cluster into smaller clusters based on nonsimilarity of data points.
An example of clustering is well explained and illustrated in Figure 1.19.
The example explains how three clusters have been formed based on the
similarity of animal species by using an unsupervised clustering technique.
(b) Association analysis: This unsupervised learning approach was intro-
duced by Rakesh Agrawal, Tomasz Imieliński, and Arun Swami. The
basic idea was to make a market basket analysis of the products sold
in a supermarket. This approach helps in finding the itemsets that are
frequently purchased together so as to understand the customer purchasing
patterns. This is done by generating more and more new rules by analyzing
the dataset. Usually, the dataset considered is large, and the final goal
of association rule learning is to train a machine to be able to mimic the
human brain for feature extraction and abstract association capabilities.
Let us understand the concept better with the help of an example. For
example, a rule {milk, coffee} ⇒ {sugar} that is discovered based on the trans-
action sales of a supermarket indicates that customers have shown a strong
tendency to buy the product sugar when milk and coffee have been bought
together. Such kind of rules generated can help a supermarket increase its
sales through better product placements and promotional pricing of items
when bought together.
If we consider Table 1.2, we find that there are seven transaction details
being considered. In each transaction, it shows what the lists of items
purchased by a customer are. Now, the interestingness of a rule can be

Table 1.2 Transaction details of a supermarket store.

TID Purchased items

1 Milk, coffee, butter

2 Eggs, butter, jam
3 Milk, coffee, sugar
4 Spices, rice, juice
5 Milk, coffee, eggs, cake, cream, sugar
6 Milk, coffee, biscuits, sugar
7 Milk, juice, soap, butter
 1.4 Components of Soft Computing 27

evaluated by using the standard constraints called support and confidence.

While support determines how frequently itemsets appear in a dataset,
confidence indicates how often a rule has been found to be true.
To understand how support is calculated for association analysis, let us
consider a rule. If a rule X ⇒ Y holds with support S, it means S% of the
transactions in the dataset contain X ∪ Y . This support value S is compared
with the threshold support value, and if S is found to be greater than the
threshold support value, it is said to have the minimum support. Again, if the
rule X ⇒ Y holds with confidence C, it means C% of the transactions in the
dataset that contain X also contains Y . This confidence value C is compared
with the threshold confidence value, and it is said to have the minimum
confidence if C is found to be greater than the threshold confidence value.
Let us proceed to find support and confidence for the rule {milk, cof-
fee} ⇒ {sugar} by considering Table 1.2. The support value for this rule will
be calculated from the given seven transactions as 3/7, which is 0.43. This
is calculated by finding out how many transactions contain the items milk,
coffee, and sugar out of a total of seven transactions. This calculated support
value is compared with the minimum support value (a threshold value)
to determine the strength of the rule. Now, if we consider the threshold
support value as 30%, the rule exceeds this value as the calculated support
value is 43%.
Another measure of interestingness of a rule is confidence. The confi-
dence value that can be calculated from the given seven transactions is 3/4,
which is 0.75. This is calculated by finding out how many transactions con-
tain all the items milk, coffee, and sugar out of a total of four transactions that
contain only items milk and coffee. This calculated confidence value is com-
pared with the minimum confidence value (a threshold value) to determine
the strength of the rule. Now, if we consider the threshold confidence value
as 50%, the rule exceeds this value as the calculated confidence value is 75%.
Based on the calculations of support and confidence, the determination of
frequent itemsets is made for market basket analysis for any departmental
store.
● Semi-supervised learning: This learning falls between SL and unsupervised
learning, and is an approach often used to deal with partially labeled data
and mostly unlabeled data. Semi-supervised learning tries to handle the
disadvantages of both supervised and unsupervised learning to solve real-time
applications. At times, data need to be hand-labeled by human, especially
by a data engineer, data scientist, or ML specialist to carry out SL. However,
this is a costly process and requires huge time for labeling all data. This is
where semi-supervised learning balances the disadvantage of SL by carrying
out training with mostly unlabeled data and only a few labeled data. Also,
28 1 Fundamentals of Soft Computing

unsupervised learning, owing to its limitations of dealing with only unlabeled

data, cannot handle a variety of applications. Semi-supervised learning also
resolves this issue, as it considers a few labeled data during the training process.
As semi-supervised learning is more effective, it is used to solve a variety of
real-world ML problems. In cases where 100% of the dataset is unlabeled, this
learning technique considers manual labeling of a small amount of data to train
the model, also called the base model. Only a small portion of data is manu-
ally labeled, as it is a very tedious and time-consuming process. The next step
is the process of pseudo-labeling that uses the partially trained base model and
makes predictions for the rest of the unlabeled data. In each iteration the unla-
beled data gets labeled, thereby increasing the number of labeled data for the
entire dataset. Such unlabeled data are separately trained, and given labels are
called pseudo-labels, as these labels will not be fully accurate. With more num-
ber of iterations, the accuracy of the model is increased, and the model is then
considered ready for any new predictions to be made.
● Reinforcement learning (RL): RL is a feedback-based ML technique that
relies on a software agent for performing some actions. The agent behaves
in an environment by performing some actions and accordingly produces
results. If the agent performs a correct action in the learning environment,
the environment in turn sends feedback to the agent in the form of an award.
However, if the agent performs an incorrect action, the environment sends a
penalty as a reward to the agent. Figure 1.21 illustrates this simple feedback
system of RL. Based on the feedback, the agent accordingly tries to adapt or
improve to the environment. One advantage of this technique is that the agent
need not be pretrained about the environment, as it can learn about it through
the feedback it receives during the training phase.
RL has recently gained much importance, as its applications have increased
over time in the last decade. The concept of RL is applied in playing games
(such as chess), robot decision-making, self-driving cars, operations research,
GA, swarm intelligence, and many more. The primary difference between SL

Actions

Reward

Learning environment Agent

Figure 1.21 Reinforcement learning.

 1.4 Components of Soft Computing 29

and RL is that SL makes a decision based on the initial input provided, whereas
RL makes a decision sequentially by considering the inputs generated at every
new phase of RL. The decisions made in RL at each level are independent of
each other, whereas the decisions made in SL rely on the decisions made at the
previous level.

1.4.5 Other Techniques of Soft Computing

There are several other soft computing techniques such as PR, chaos theory, and
evidential reasoning (ER). A brief description of all of these techniques is given
next:
● Probabilistic reasoning (PR): Real-world situations many a times deal with
uncertainty. In such cases, PR may be used to indicate the degree of uncertainty
involved. As we know, probability is the chance that an uncertain event will
occur, and its value is between 0 and 1. PR uses this concept of probability to han-
dle uncertain situations. Generally, the probability of occurrence of an uncertain
event is calculated as:
number of desired Outcomes (N)
Probability of occurrence (P) =
total number of outcomes (T)
A belief network, also known as a Bayesian network or Bayes network, is
a probabilistic graphical model used to represent a set of variables and their
conditional dependencies via a directed acyclic graph (DAG). The nodes in a DAG
represent variables (can be unknown parameters, latent variables, observable
quantities, etc.), and the edges represent conditional dependencies. If two nodes
are not connected in the graph, it indicates that the two variables are conditionally
independent of each other. A belief can be considered as the occurrence of a
state. For instance, a switch can have two states – either on or off . The probability
of the occurrence of the state {switch = “on”} is 50%. Similarly, the belief in the
state {switch = “off”} is also 50%. In this way, for a given problem, all beliefs of all
possible states of a node are stored in a table called Conditional Probability Table.
To define conditional probability using a traditional approach, the concept of
joint probability is used. For instance, for an uncertain event Q, the probability
P(Q|R) represents a belief in P, assuming that R is known. The conditional
probability formula P(Q|R) can be equated using the concept of joint probability
P(Q, R) as given below.
P(Q, R)
P(Q|R) =
P(R)
However, it is also possible to define P(Q | R) without referring to the joint prob-
ability as follows:
P(R|Q)P(Q)
P(Q|R) =
P(R)
30 1 Fundamentals of Soft Computing

This is called Bayes rule and is often used for updating a belief about a hypothesis
Q in the light of new evidence R.
Bayesian networks provide a natural representation of (casually induced)
conditional independence. It has replaced many rule-based systems by being less
rigorous and more robust. However, one major limitation of a Bayesian network
is that it typically requires initial knowledge of many probabilities. Another issue
is that the output is also largely dependent on the quality and extent of prior
knowledge.

● Chaos theory: Chaos theory is well-suited to be used for problems that are
highly sensitive to initial conditions. In such a case, a slight difference in ini-
tial conditions (for example, a change in initial value in the second place after
the decimal point) leads to highly diverging outcome. The chaotic behavior can
be experienced in our nature, such as changes in weather. Robert L. Devaney has
classified a dynamic system as chaotic based on the following three properties:
– It must be sensitive to initial conditions (the “butterfly effect”): The data points
in a chaotic system are arbitrarily in close approximation to each other with
significantly different future paths.
– It must be topologically mixing: The topological transitivity or topological mix-
ing relates to the evolution of the system over time such that any given region
may eventually overlap with another region.
– It must have dense periodic orbits: Every point in the space is approached arbi-
trarily closely by periodic orbits.
To sum up, chaos theory as defined by Kellert is “the qualitative study of
unstable aperiodic behavior in deterministic nonlinear systems” (Kellert 1993,
p. 2). As understood, a chaotic system is nonlinear and sensitive to initial
conditions. There is also no periodic behavior in such systems, and the motion
remains random. Considering these characteristics, a few applications based on
chaos theory include observation of weather patterns, stock market predictions,
algorithmic trading, bird migration patterns, observation of social phenomena,
robotics, and study of brain waves.
● Evidential reasoning: Evidential reasoning (ER) is a recent approach that
has been developed mainly on the basis of AI, decision theory, statistical
analysis, and computer technology. In decision theory, ER approach is a
generic evidence-based multi criteria decision analysis (MCDA) approach
and can deal with problems having uncertainties that include randomness
and ignorance. ER supports assessments, decision analysis, and evaluation
activities. For instance, ER contributes to the environmental impact assessment
for developing a proposal or a project related to the environment.
The ER approach constitutes the modeling framework of multi-criteria
decision-making (MCDM) problems using the following concepts:
 Exercises 31

⚬ Belief structures: The belief structure is used to represent the performance

of an alternative assessed against a criterion.
⚬ Belief decision matrix: A belief decision matrix is a list of values stored in
rows and columns in which each element is a belief structure and is used to
identify, analyze, and rate the performance of relationships between sets of
values.
⚬ ER algorithms: This algorithm is mainly governed by using four basic steps:
generation of basic probability masses, combining basic probability masses,
generation of combined degrees of belief, and generation of utility interval.
⚬ Rule- and utility-based information transformation techniques:
It involves both qualitative and quantitative data for decision-making to
transform various sets of assessment grades to a common framework of
discernment.
As soft computing is tolerant of uncertainty, imprecision, and approximation,
it allows researchers to carry out various experiments to solve real-world complex
problems. As discussed at the beginning of this chapter, the three core components
of soft computing include the fuzzy logic-based computing, neurocomputing, and
GA. All these techniques contribute greatly to develop probable solutions to a
problem rather than a precise fixed solution. Hence, soft computing is a boon to
the computing world for solving problems that are not possible to be solved by
traditional computational models.

Exercises

A) Choose the correct answer from among the alternatives given:

a) The truth values of fuzzy logic can be:
i) Either 0 or 1
ii) Less than 0 or greater than 1
iii) Between 0 and 1
iv) Between −1 and 1
b) Which among the following is not a part of a neuron?
i) Dendrite
ii) Axon
iii) Soma
iv) Spinase
c) Which among the following is not a regression technique?
i) Polynomial regression
ii) Associative regression
iii) Lasso regression
iv) Ridge regression
32 1 Fundamentals of Soft Computing

d) Each iteration of the genetic algorithm is called:

i) Generation
ii) Run
iii) Itemset
iv) Mutation
e) The probability P(A|B) in Bayes theorem can be expressed as:
P(A,B)P(B)
i) P(A|B) = P(A)
P(B|A)P(A)
ii) P(A|B) = P(B)
P(B|A)P(B)
iii) P(A|B) = P(A)
P(A,B)P(A)
iv) P(A|B) = P(B)
f) In association analysis, confidence ___________
i) Determines the total number of itemsets in a dataset
ii) Determines how frequently itemsets appear in a dataset
iii) Indicates how often a rule has been found to be true
iv) Indicates how many itemsets are purchased on an average per transac-
tion
g) Each connection link in ANN is linked with __________ that contains stat-
ics about the input signal.
i) Activation function
ii) Bias
iii) Weights
iv) Neurons
h) Where can the Bayes rule be used?
i) For answering probabilistic queries
ii) For dealing with training and testing data
iii) For building a decision tree
iv) For creating clusters of nodes
B) Answer the following questions:
1) Define soft computing. Explain, in brief, the various components of soft
computing.
2) Who is the founder of fuzzy logic? How is a membership function defined
in fuzzy logic?
3) Explain the following machine learning techniques:
a) Association analysis
b) Clustering
c) Reinforcement learning
4) Differentiate between:
a) Soft computing and hard computing
b) Crisp logic and fuzzy logic
 Exercises 33

c) Supervised learning and unsupervised learning

d) Classification and regression
e) Probabilistic reasoning (PR) and evidential reasoning
5) What is an artificial neuron? List some commercial practical applications
of ANNs.
6) Define the terms chromosome and crossover as used in genetic algorithms.
Explain, in detail, all the various phases of genetic algorithms.
7) State the importance of semi-supervised learning. List some commercial
practical applications of machine learning.
8) Explain the various types of machine learning. What are the applications
of supervised machine learning in modern businesses?
Random documents with unrelated
content Scribd suggests to you:
Strúma, a wen or swelling in the necke or arme-pits, wherin be as it were
hard kernels closed in the skin. Some haue also taken it to be the Kings
euill. Vsed also for a bunch in the backe.

Strumático, that hath a Strúnia.

Strumẻa, some take it for the hearbe Crowfoot which they say cureth the
Kings euill, others take it for a kind of hearbe that beggers vse to make
their flesh raw.
Strumẻntále, as Stromẻntále.
Strumẻnto, as Stromẻnto.
Strúmo, a plaister made of Strumẻa.
Strumóso, as Strumático. Vsed also for one that hath a great throate.
Strupiáre, as Stupráre.
Strupiatóre, as Stupratóre.
Strúpo, proud, haughty, disdainefull, without reason. Also a Stúpro.
Strusciáre, as Strutiáre.
Strúscio, as Strútio.
Strutẻa, a kind of Peare-quince. Also a kind of Quinces.
Struthopódi, a kinde of people hauing feet like sparrowes.
Strutiáre, to excruciat, to vex, to tug, to hauocke.
Strútio, a vexing or excruciating. Also a bird called an Estridge or Ostrich.
Strutióne, Sope-weed, Sope-wort or Fullers-hearbe.
Struttíbile, that may be melted, thawed, consumed or destroied. Also
made of diuers pieces or diuers things.
Struttióne, vsed for destruction, as Struttúra.
Strútto, destroied, consumed. Also melted or thawed. Also built, compact,
or framed. Also instructed. Also construction or building. Also sewet or
tried butter.
Struttóre, a builder, a framer, a Mason, a Carpenter. Also a melter or
founder.
Struttúra, composition, structure, framing, compacting together as a
building.
Strúzza, a kind of hunting with hounds.
Struzzicáre, to make cleane, to trim vp finely. Also as Stuzzicáre.
Struzziére, an Estridge-keeper.
Strúzzo, an Estridge or Ostrage.
Struzzóso, a yongue Estridge.
Stù, vsed for Se tù, if thou.
Stúa, as Stúfa.
Stuáre, as Stufáre.
Stuccáre, to parget, to daube or worke with Stúcco, lome or morter. Also
to stanch or stop. Also to drie, to stiffen or to harden. Also to glut, to
gorge, to cloy, to satiate.
Stuccatóre, a worker in Stúcco, or of images.
Stúcce, a kinde of file that Goldsmithes vse.
Stucchéuole, tedious, wearisome, importunate. Also wallowish in taste.
Stucchézza, satiety, fulnesse, glutting.

Stúcchio, an estwefe, a pocket cace with Cizers, Penkniues, &c.

Stucciétto, as Stúcchio.
Stúccio, as Stúcchio.
Stúcco, a kinde of pargeting stuffe much vsed in Italy to build statue or
imagery worke, composed of paper, sand and lime with other mixtures,
the imagery worke in the inner court at Nonesuch in England is made of
such. Also glutted, satiate, gorged, cloied. Also stanched or stopt. Also
stifned or hardned.
Studẻnte, as Studiánte.
Studiánte, a student, one that studieth.
Studiantúzzo, a simple student.
Studiáre, to studdy, to endeuour or apply the minde, to giue ones minde
vnto.
Studiáre il piéde, to heede ones foote, to run away.
Studiétto, a little Studdy, Cabbanet or Closet.
Stúdio, an earnest bending of the minde to a thing, affection to doe good
or euill, study, endeuour, diligence, industry, laborious desire, mentall
exercise.
Stúdio, a priuate study, cabinet, closet or any place to study in. Also an
vniuersity where studies are profest. Also a Colledge where Students
 are. Also a standing deske in a schoole for great bookes.
Studiólo, as Studiétto.
Studiosaménte, studiously, with diligence and affection, desirously,
carefully. Also aduisedly, of set purpose, for the nonce, expresly.
Studiosità, studiousnesse.
Studióso, studious, diligent, earnest, carefull, musing, desirous of learning.
Stúfa, a stoue, a hot-house.
Stufaiuólo, a stoue or hothouse keeper.
Stufáre, to stoue close. Also to bath in a hot-house or stoue.
Stufaruólo, as Stufaiuólo.
Stufátto, Vsed for Stupefátto.
Stufẻllo, a pipe, a fluite, a recorder.
Stufétta, a little stoue or hot-house.
Stuffáre, to glut, to cloy, to satiate. Also as Stufáre.
Stuffáta, a pot of stewed meate.
Stufféuole, as Stucchéuole.
Stúffo, gorged, glutted, cloid, satiate.
Stúfo, the headinesse or fumosity of wine or any thing else.
Stuiẻllo, a tent to tent a sore with.
Stúmia, scum, fome, froth.
Stumiáre, to scum, to fome, to froth.
Stumióso, scummy, foamy, frothy.
Stuóia, as Stóia, any mat.
Stuoiáre, as Stoiáre, to mat.
Stuoiáro, a Matter, a Mat-maker.

Stuóla, as Stóla.
Stuoláre, as Stoláre.
Stuólo, a troope, a band, a squadron, a rout or company of men. Also a
knot or crew of good fellowes, a beauy of Ladies, a flocke of faire
women. Also a shoale of fishes, a flight of birdes, a swarme of Bees, &c.
Stuóra, as Stóia, a Mat.
Stuoráre, as Stoiáre, to mat.
Stuoráro, as Stoiáro.
Stupefáre, fáccio, féci, fátto, to astonish or dismay, to become abashed, to
driue into a dumpe, to be past all sence, to wonder and gaze at with
admiration. Also to be amazed or astonished. Also to stupifie.
Stupefattióne, stupefaction, amazement, astonishment, wonderment.
Stupefátto, astonished, dismaied, abashed, in a dump, past all sence, &c.
stupified. Looke Stupefáre.
Stupẻndo, admirable, wonderous, maruellous, working admiration.
Stupidézza, as Stupidità.
Stupidíre, dísco, díto, as Stupefáre.
Stupidità, stupidity, dulnesse, vnsensiblenesse, suddaine priuation of
sences, amazement, wonderment, astonishment, a benumming of the
minde.
Stúpido, stupide, sencelesse, blockish, dull and benummed in sences. Also
fearefull, abashed, amazed, astonished, dismaied, foolish.
Stupíre, písco, píto, to wonder at with admiration.
Stupíto, wondered at with admiration.
Stuporatióne, as Stupidità.
Stupóre, wonderment, amazement, suddaine lacke of sence or feeling.
Stuporóso, full of wonderment.
Stupráre, to rape, to rauish or deflower a virgin. Also to commit rape or
adulterie.
Stupratióne, a rauishing.
Stupratóre, a rauisher or deflowrer of women. Also an adulterer.
Stúpro, rape, rauishment, deflowring of women, adultery, whoredome.
Stupróso, adulterous, full of or giuen to whoredome, adultery or
rauishment.
Sturáre, to vnstop, to open, to vent.
Sturbaménto, a disturbance, a troubling.
Sturbáre, to disturbe, to trouble, to molest.
Sturbatióne, as Sturbaménto.
Stúrbo, trouble, disturbance, vexation.
Sturióne, the fish called Sturgion.
Sturmẻnto. Vsed for Stromẻnto.
Stúrno, as Stórno.

Stúrzo, an Estridge bird.

Stutáre, to doe out, to quench forth.
Stúua, as Stúfa.
Stuuaiuólo, as Stufaiuólo.
Stuccáre, as Stufáre.
Stuzzicadẻnti, a tooth-picke.
Stuzzicaménti, prouocations, allurements, ticklings, enticements,
dalliances.
Stuzzicaorécchi, an Eare-picker.
Stuzzicáre, to prouoke, to entice, to allure, to tickle vnto, to egge or pricke
on, to fumble, to grope, to play or dally withall wantonly.
Stuzzicatóre, a prouoker, an enticer, a tickler, an egger or pricker on. Also
a fumbler or a groper.
Stúzzico, a prouocation, an allurement, a tickling, a pricking or setting on.
Sù, vp, vpward. Also on ouer or vpon. Also an aduerbe of encouragement
vp and away, forward, on, goe to.
Súa, hirs, hir owne. Also his or his owne, for note that the Italians make
euer the six Pronounes possessiues to agree with the thing spoken of
and not the person.
Suáda, the Goddesse of Eloquence or delectable speech.
Suadáta, open, vnkept, free for all men.
Suadẻla, perswasian, faire speech.
Suadére, suádo, suási, suáso, to perswade or aduise vnto with faire
words, to induce, to moue or exhort one to be of our opinion.
Suadíbile, that may be perswaded.
Suádo, perswaded with faire words.
Suagáre, to stray, to roaue or wander a side.
Suaghíre, ghísco, ghíto, the contrary of Inuaghíre, to fall out of liking with.
Suagolonáre, to gad, to loiter or roaue vp and downe.
Suagolóne, an idle gadder, a loitering companion.
Sualiáta, garishly deckt, gallantly trimmed.
Sualigiaménto, a robbing, an vncasing, a stripping or rifling of one.
Sualigiáre, to rob, to strip, to fripe, to rifle or vncase one by the high-way.
Sualisciáre, as Sualigiáre,
Sualláre, to vnually, to come or fetch or driue out of a vally.
Sualtẻrnico, a kind of Amber.
Suampáre, as Vampáre, to flash.
Suangẻlizzáre, as Euangẻlizzáre.
Suaniménto, a swouning. Also a vanishing out of sight.
Suaníre, nísco, níto, to vanish out of sight. Also to fade away. Also to
droop or faint.
Suaníto, vanished out of sight, faded away, drooped or fainted. Also wine
or any drinke that is flat or dead, and that hath lost his strength.

Suantaggiáre, to disuantage.
Suantággio, disaduantage, oddes.
Suantaggióso, disuantageous.
Suaporaménto, a breathing forth.
Suaporáre, to euaporate or breath forth.
Suariánza, variation, diuersity.
Suariáre, to vary, as Zauariáre.
Suariataménte, variously.
Suariáto, varied. Also motly coloured.
Suário, of or belonging to swine.
Suarióne, a varier, as Rauarióne.
Suáro, a fish called in Latin Lacertus.
Suasíbile, that may be perswaded.
Suasióne, perswasion, exhortation.
Suáso, perswaded or induced vnto with faire words. Also a kinde of
whitish pale colour that painters vse, some take it for a kind of sad gray.
Suasóre, a perswader with faire words.
Suasório, of or belonging to perswasion.
Suasório, perswasiue, alluring vnto.
Suáue, as Soáue, sweet, pleasant.
Suauità, sweetnesse, as Soauità.
Suázzo, a fish called a Turbut.
Suò. Vsed of Dante for Sótto, vnder, and much vsed in composed words.
Subagiráre, to wind vnder.
Subagitatióne, an vnder agitation or solliciting.
Subaltẻrnare, to succeed in order or by turne.
Subaltẻrnatióne, a succeeding in due order or by turne, subalternation.
Subaltẻrno, subalternall, succeeding in due order and by turne.
Subáre, to grunt as the sow doth for the boare to haue him doe his kind.
Subatióne, the appetite or stirring to generation in swine, the grunting of
the sow for the boare.
Subáto, as Subatióne.
Subátto, subdued or forced.
Subattríce, a lustfull sow or woman.
Subbarcáre, to bow vnder as a bow, to vnder arch, to vnder vault.
Subbẻntráre, to enter or come in vnder.
Súbbia, a Shooemakers awle.
Subbiáre, to pricke with an awle. Also to whistle.
Súbbio, a whistle. Also a Weauers yarne-beame. Also axle-tree.
Subbioláre, to rowle vpon a yarne-beame.
Súbbiolo, a little whistle. Also a little yarne-beame.
Subbísso, as it were Sottoabísso, a bottomlesse pit or gulfe, a dungeon of
hell.
Subbolíre, as Sobbolíre.
Subcelẻste, of vnder heauen.

Subdiuídere, vído, vísi, víso, to subdeuide.

Subdiuisióne, a subdiuision.
Subdólo, full of deceit, craft or wiles.
Subdolóso, as Subdólo.
Súbe, a kind of vnluckie Skreech-owle which is said to breake all the
Eagles egges he can come vnto.
Subẻntráre, to enter or come in vnder.
Súbero, the tree or wood Corke.
Súbet, a kinde of disease in a horse.
Subgiontióne, a subiunction.
Súbia, a Shoemakers awle.
Subiáre, to pricke with an awle.
Subiẻttiuaménte, subiectiuely.
Subiẻtto, subiect. Also a subiect.
Subilláre, by meere importunity and earnestnesse, and as it were against
ones will to entreate and obtaine a suite of a man.
Subintẻllígere, to meane or vnderstand.
Subintẻndere, to meane or vnderstand.
Súbio, as Súbbio.
Subíre, bísco, bíto, to grunt as the sow doth for the bore to doe his kinde
to long or lust for most beastly.
Subissáre, as Precipitáre.
Subitaménte, suddenly, one a sudden.
Subitána, suddaine, Mórte subitána.
Subitáneo, suddaine, vnthought vpon.
Subitézza, suddainenesse, hastinesse, speed.
Súbito, suddaine, suddenly, vnlooked for.
Súbito chè, so soone as, presently vpon.
Súbito cóme, idem.
Subitóso, suddaine, hasty, full of speed.
Sublimáre, to raise, to extoll, to aduance, to set on high. Also to
sublimate.
Sublimatióne, sublimation, subliming, lifting vp.
Sublimáto, high raised, extolled, sublimed, sublimated, that is aboue vs.
Sublimáto, sublimate, as Solimáto.
Sublíme, sublime, high and aloft, extolled, aboue vs lofty. Also the vpper
post of a doore. Also the mantle of a chimny.
Sublimità, sublimity, highnesse, the top or tip of any thing, height,
loftinesse.
Sublínguio, the wezell of the throate.
Sublitióne, the ground colour whereon other coloures are laid, Painters call
it grasing.
Sublunáre, whatsoeuer is, breedeth or liueth vnder the Moone.
Subluogáre, to lay, to place or set vnderneath. Also to dispose vnder hand.

Sublúuie, any filth or ordure. Also a disease in Cattell betweene the

clawes.
Subodoráre, to smell vnder hand, to haue an inkling or notice of any thing.
Subóido, moist, humide, wettish.
Súbole, as Sóbole.
Subordinánza, a subordination.
Subordináre, to subordinate.
Subordinatióne, subordination.
Subornaménto, as Subornatióne.
Subornáre, to suborne, to debauch, to intice, to misleade to euill or to
beare false witnesse, to instruct and send one priuily to doe or say any
thing to the end to deceiue.
Subornatióne, subornation, misleading, enticing or debauching vnto some
euill.
Subornatóre, a suborner, or misleader, an inticer or debaucher vnto euill.
Subrogáre, as Surrogáre.
Subrogatióne, as Surrogatióne.
Substantíuo, that may stand of it selfe, a Substantive.
Subsedére, séggo, sedéi, sedúto, to vnder sit.
Subsistẻnte, subsisting.
Subsistẻntia, a subsistance or ground-worke of any thing, an abiding.
Subsístere, sísto, sistéi, sistúto, to subsist, to abide still, not to recoyle or
giue place. Also to resist or withstand. Also to substitute.
Subsistúto, subsisted. Also substituted or appointed. Also a substitute.
Subsoláno, the East or Easterly wind.
Subtẻrfúgio, as Sottẻrfúgio.
Subucáre, to lurke or sneake in a corner.
Subuẻntióne, helpe, supply, succour.
Subuẻrsióne, as Souuẻrsióne.
Subugliáre, as Sobugliáre.
Subúglio, as Sobúglio.
Subulóne, a Deare called a Spitter. Also a yongue foole.
Suburbanità, neighbour-hood of them that dwell in a City.
Suburbáno, Lat. neere vnto or belonging to a Citty.
Sù carpíre, to snatch vp.
Succedána, any drug that may be vsed for default of another,
Apothecaries call it quid pro quo.
Succedáneo, that succeedeth or commeth in place of another.
Succédere, cédo, cẻssi, cẻsso, or cedúto, to succeed, to follow, to haue
good successe, to prosper and come to passe, to ensue in the place of
another.
Succedúto, succeeded, followed. Also come in the place of another,
ensued.

Succeneríto, baked vnder the ashes.

Succẻssíbili béni, goods that come by succession, successible goods.
Succẻssióne, succession, a succeeding, a comming in place of another.
Succẻssiuaménte, successiuely.
Succẻssíuo, successiue, succeeding.
Succẻsso, succeeded. Also a successe or comming to passe.
Succẻssóre, a successor, a succeeder, one that commeth in place of
another.
Succhiacápra, a birde which is said to sucke a Goates vdder, some call it a
winde-fucker.
Succhiáre, to sucke. Also to like or smacke.
Succhiasángue, a Horse-leech, a Bloud-sucker.
Succhiaspína, a louer of wine, a licke or sucke spigot.
Succhiatóre, a sucker, a sucklin.
Succhiauígna, as Succhiaspína.
Succhiẻlláre, to spigot or boare thorow.
Succhiẻllo, an auger, a piercer, a wimblet, a gimblet, a boarer, a forcebead.
Also a spigot. Also a little pipe or flut. Also a bird windfucker.
Súcchio, the iuice of any thing, the sap or moistnesse of trees or any thing
else. Also moistnesse or waterishnesse. Also sucking or smacking till a
print remaine, Veníre in Súcchio, to haue ones teeth run a water with
longing earnestly for any thing.
Succhióse lábbia, sucking-plumbe lips.
Súcci, mi páre ródere bombétti & súcci.
Succiaménto, any sucking or smacking.
Succiáre, as Succhiáre, to sucke.
Succiatóre, a sucker. Also some part of a pump.
Súccida, vnwashed or new shorne-wooll.
Succídere, cído, císi, císo, to cut vnder.
Succidézza, as Succidúme.
Succidíre, dísco, díto, to foule, to sully or defile with grease or sweat.
Súccido, filthy, foule, sluttish, slullied, greasie, sweaty, vnwashed.
Succidúme, all manner of filth, filthinesse, sluttishnesse, or sweaty grease.
Succináre, to make a low or soft noise. Also to sing a base or drone tenor.
Succínco, of or belonging to Amber.
Succingẻnte, vnder-guirding.
Succíngere, to vnderguirt or cingle. Also to enuiron, to compasse or fence
about.
Succíno, gum or rozin of sweet-trees, namely of the Pine-tree, which by
reason of the clearenesse thereof was in old times called glasse. The
first and best kind is yellow and bright Amber, the next is white and
duskish called, bastard-Amber. Also of or belonging to Amber. Also a
kind of Plum or Damzen-tree. Also as Lincúrio.

Succintaménte, succinctly, briefly.

Succínto, succinct, compendious, short, materiall, pithie. Also trussed vp
or vnder-guirt. Also enuironed.
Súccio, a sucke, a sucking, a smacke. Also quickly, in a moment, instantly.
Súcciola, a sucking. Also a sucking babie.
Succioláre, to sucke. Also to smacke.
Súcciole, greene Chesnuts sodden in water.
Succipiéno, so full that it may be suckt.
Succíre, císco, cíto, to sucke, to smacke.
Succísa, the hearbe Diuels-bit.
Succíso, cut, sliced, gasht or hewed vnder.
Súcco, the iuice, the sap or moistnesse of any thing. Vsed also for vigor or
strength. Also a kind of iron instrument to picke lockes with.
Succobrináre, to play the Ape or Munkie and imitate any gesture.
Succobríno, a mimike gesturer, that counterfeites all gestures as a Munkie.
Succómbere, as Soccómbere.
Succóso, moist, full of iuice or sap.
Succuba, a whore lying with another womans husband.
Súccubi, as Succúmbij.
Succúmbij, certaine spirits that take the shape of women vpon them, and
so haue the companie of men.
Succussário, a slow, dull, heauie-going iade.
Succussóni, slow, dull or heauie iades.
Sucína, a kind of Plum or Damzen.
Sucíno, a kind of Plum or Damzen-tree.
Súco, as Súcco, iuice or sap.
Sucóso, full of iuice or sap, moist.
Sucúbij, as Succúmbij.
Súcula, the middle poste of any winding crane or wind-mill to draw vp
weights.
Súcule, the seauen Starres appearing in Iune called the Hiades.
Sud, the South-winde.
Sudáccio, filthy stinking sweet.
Sù dal ciẻlo, vp from heauen.
Sudaménto, any kind of sweating.
Sudámini, red angry pimples.
Sudánte, sweating, dropping with.
Sudarcáto, a sicknesse called the Iaundise.
Sudár di méle ẻt látte, to sweat or flow with milke and honnie.
Sudáre, to sweat, to drop with.
Sudário, a wiper, a towell, a rubber, a handkercher to wipe the sweat
away.

Sudatíccio, sweatie, enclined to sweat.

Sudatióne, any sweating.
Sudatório, a hot-house or sweating place.
Sudatório bágno, a hot or sweating-bathe.
Suddiacóno, a Subdeacon, a Subdeane.
Suddíccio, slouenlinesse, filthinesse.
Suddiccióne, a filthy, slattish, slouenly fellow.
Suddiccíre, císco, cíto, as Sudiccíre.
Súddito, a subiect, a vassall, an vnderling.
Sudest, the South-east winde.
Sudétto, aboue said, or forenamed.
Súdi, a kind of excellent good Fish.
Sudiccíre, as Succidíre.
Súdice, as Súccido.
Sudiciménto, seducement.
Sudiciótta, a filthy foule slut.
Sudiciúme, as Succidúme.
Súdo, a great fish as Sphiréno. Also the cleare firmament, or faire and dry
weather after wet and raine.
Sudóre, sweat. Also labour or trauell.
Súe, hirs, his, or theirs. Also as Sù.
Suecchiáre, to grow, to become from old yongue againe.
Suecchiáto, become yongue againe, refreshed with youth or vigor.
Suefáre, to bring in vse or custome.
Suefattióne, vse, vre, custome.
Suefátto, accustomed, enured, wont.
Suegghiáre, as Suegliáre.
Suegghiaruólo, as Suegliaruólo.
Suegghiáto, as Suegliáto.
Suegghiatóio, as Suegliaruólo.
Sueggiáre, as Suegliáre.
Suéglia, a kind of racke or torture in Florence. Also a hautboy or such
winde instrument.
Suegliáre, to awaken, to wake, to watch.
Suegliaruólo, a waker, a larum in a clocke.
Suegliáto, awakned, waked, watcht.
Suegliáto, watchfull, vigilant.
Suegliatóio, as Suegliaruólo.
Suegliatóre, as Suegliaruólo.
Suẻgliere, suẻlgo, suẻlsi, suẻlto, as Suẻllere.
Sueglióne, a hautboy or loud instrument. Also the drone of a bagpipe.
Sueláre, to vnuaile, to vn ouer, to disclose.
Suelenáre, as Sueleníre.
Sueleníre, nísco, níto, to vnpoison, to vnrankle.
Suẻlgere, as Suẻllere.
Suẻllere, suẻlgo, suẻlsi, suẻlto, to extirp, to grub or root out, to pull vp by
the rootes.
Suẻlta, a plucking or rooting out.
Suẻlto, a place from out which all things are pulled. Also loose of the root,
extirped, grubbed vp or rooted out.

Suelutáre, to make bare, to vnueluet.

Suembráre, as Smembráre.
Suemoráre, as Smemoráre.
Suenáre, to cut the veines of ones throat.
Suenáta bríglia, a bit with an open mouth as ports or vpsets.
Suenatóre, a murtherer, a killer, a butcher.
Suenatúra, Riders call it open mouthes, as a Catsfoote or other port or
vpset.
Suenéuole, vncomely, vnseemely.
Sueniménto, a swouning or fainting.
Sueníre, to faint or swoune.
Suentáre, to euaporate, to blow or vent out, to let the winde, breath,
spirit, force or wit of any man or thing goe out. Also to ayre in the
winde.
Suentáta, an idle, a gadding or humorous toy, fancie or conceit.
Suentáto, euaporated, vented or breathed out. Also ayred in the winde.
Also a light-headed fellow.
Suentoláre, to waue in the ayre or wind. Also as Suentáre.
Suéntolo, any fan to make winde with.
Suentráglia, the guts, offall, garbish, panch or entrailes of any creature.
Suentráre, to vnpanch, to vnbowell.
Suentúra, a misfortune or mischance.
Suenturáto, vnfortunate, vnluckie.
Suenturóso, vnfortunate, vnhappie.
Suenúto, fainted or fallen into a swoune.
Suerginaménto, a rape, a deflouring.
Suergináre, to defloure, to rauish.
Suergináta, a rauished or defloured wench.
Suergógna, shame, reproch, impudencie.
Suergognáre, to shame, to reproch.
Suergognáto, shamed, shamelesse, impudent.
Suẻrnánte, wintring, passing the winter.
Suẻrnáre, to wax Sommer. Also to winter or to passe the winter. Also vsed
of Dánte for the chirping of birds, because winter is done, to vnwinter.
Suẻrre, as Suẻllere.
Suẻrtáre, to tell or vtter whatsoeuer one knowes, to be Tom-tell-troth.
Suẻrza, a splinter or a shiuer.
Suẻrzáre, to splinter, to shiuer.
Suẻrzáto, burst, cleft, cloued, riuen.
Suẻstíre, suẻsto, suẻstíj, suẻstíto, to disaray, to disattire, to vncloth, to
strip naked.
Suẻstíto, disaraied, disrobed, vnclothed, disattired, stript naked, vnclad.
Suéta, a fish called in Latine Nasus.
Suetíno, a kind of yellow precious stone.
Suẻto, accustomed, wont, enured.
Suettáre, to cut off or lop off the top of any flower or plant.
Suetúdine, wont, custome, vre.
Suezzáto, disaccustomed, vnwonted.
Suffarcináre, to stuffe full, or fill vp. Also to trusse vp. Also to lode or
burthen.

Sufficiẻnte, sufficient, enough, able.

Sufficiẻnteménte, sufficiently.
Sufficiẻnza, sufficiencie, abilitie.
Suffígine, a sweet perfume.
Suffiláre, to whistle, to whizze.
Suffílo, a kind of disease in the eares.
Suffiménto, as Suffumicatióne.
Suffocáre, as Soffocáre, to stifle.
Suffocatióne, a suffocation, a stifling.
Suffocatóre, a strangler, a stifler.
Suffogáre, as Soffocáre.
Suffoláre, to whistle, to hizze, to whizze.
Suffólcere, as Soffolcíre.
Suffolétto, a little whistle.
Suffolíno, a little whistle.
Súffolo, any kind of whistle or hisse.
Suffólto, as Soffolciúto.
Sufformatíuo, sufforming.
Suffragáno, as Soffragáno.
Suffragáre, as Soffragáre.
Suffragatióne, as Soffragatióne.
Suffragatóre, as Soffragatóre.
Suffragatório, as Soffragatório.
Suffraginóso, as Soffraginóso.
Suffrágio, as Soffrágio.
Suffrágo, as Soffrágo.
Suffrenáre, as Soffrenáre.
Suffrenáta, as Soffrenáta.
Suffrenatióne, as Soffrenatióne.
Suffríggere, as Soffríggere.
Suffrítto, as Soffrítto.
Suffúlcere, as Soffolcíre.
Suffuléno, a Bull-finch.
Suffumicáre, to vse suffumications, to fume or perfume vnderneath. Also
to burne Incense or Sacrifice. Also to smoake or dry in the smoake.
Suffumicatióne, a suffumication, a fuming or perfuming from vnder. Also a
burning of Sacrifice.
Suffumígij, as Suffumicatióne.
Suffusióne, a suffusion, a spreading abroad or vpon. Also a disease in the
eyes called a pin and a web.
Sugáia cárta, paper that suckes inke.
Sugamáno, a hand-towell or wiper.
Sugáre, to wipe and dry, as Sciugáre. Also to sucke vp.
Súgaro, the tree or wood called Corke.
Sugatóio, a hand-towell or wiper.
Sugẻlláti, blacke and blew spots of dead bruised bloud vnder the skin.
Súgero, the tree or wood Corke.
Suggẻlláre, to seale, to signe, to stampe.
Suggẻllatióne, bloud congealed vnder the tunicle of the eyes, lying there.
Suggẻllo, a seale, a signet, a stamp.
Súggere, súggo, suggéi, suggiúto, or sútto, to sucke dry, to sucke out or
from.
Suggeríre, rísco, ríto, to suggest vnto or put in minde, to shew and tell
withall. Also to minister, to finde or allow vnto. Also to yeeld and send
forth things whereof are plentie.

Suggẻstáre, to suggest or put in minde, to shew and tell withall.

Suggẻstióne, suggestion or putting in minde.
Suggẻsto, suggested or put in minde, shewed and told withall. Also
ministred or allowed vnto.
Suggẻsto, a high chaire, a pulpit or chaire whereout Orations or Sermons
are made. Also a high bush or tuff of haire which ancient Matrons were
wont to rowle somewhat high vpon their heads.
Suggẻttáre, to subiect, to bring vnder or in subiection, to submit.
Suggẻttéuole, that may be subiected or brought vnder.
Suggẻttióne, subiection, seruitude, thraldome, bondage.
Suggẻttíre, tísco, títo, as Suggẻttáre.
Suggẻttíto, subiected, brought vnder.
Suggẻttíuo, subiectiue, to be subiected.
Suggẻtto, a subiect, a vassall, an vnderling, a thrall, a bondman. Also a
subiect, a ground or theame to write, to speake or worke vpon. Also
subiected, brought vnder or enthralled. Also addicted, enclined or
subiect vnto.
Suggiacére, as Soggiacére.
Suggiugáre, as Soggiogáre.
Suggiúgnere, as Soggióngere.
Sugherẻllo, an Elder-pipe. Also a little Súghero.
Súghero, the tree or wood called Corke. Also a deuise made of bladders or
pieces of Corke fastned to ones shoulders to learne to swim with.
Sugliárda, as Sogliárda.
Sugliárdo, as Sogliárdo.
Súgna, as Sóngia.
Sugnáccio, filthy grease or tallow.
Sugnáre, as Songiáre, to tallow.
Sugnóso, as Songióso, tallowish.
Súgo, any iuice, sap, or moisture.
Sugóso, full of iuice, sap or moisture.
Súgulo, a kind of new must or sweet wine.
Suguméra.
Suiaménto, a misleading or straying out of the way. Also a debauching or
allurement.
Suiáre, to mislead or draw out of the right way, to debauch or entice to
euill.
Suiáto, misled, straied or drawen out of the right way. Also suborned,
enticed or debauched to euill. Also an vnthriftie, misled or debauched
man.
Suigoríre, rísco, ríto, to weaken, to depriue of force, vigor or vertue.
Suillaneggiáre, to chide, to brawle or scold at. Also to offer wrong or
villanie vnto.
Suilluppaménto, as Suillúppo.
Suilluppáre, to vnwrap, to disintangle, o vnfold, to rid free.

Suillúppo, an vnwrapping, a disintangling, a free riddance, an vnfolding.

Suináre, to leake out as wine doth.
Suinchiáre, to whiske with a riding wand, our Riders say to switch.
Suinchiáta, a whiske or switching giuen with a riding wand.
Suincoláre, to free from bonds or snares. Looke Víncolo.
Suisáre, to vnface, to disuisage. Also to slash or gash ouer the face. Also
to neglect and not consider vpon.
Suisceráre, as Suentráre, to disinbowel.
Suiscerataménte, with all a mans heart and bowels, most-most hartely.
Suisceráto, vnpanched, vnboweled.
Suisceráto seruitóre, a most heartie, louing and affectionate seruant.
Suíssimo, most his, his most-most.
Suitiáre, as Disuitiáre.
Suiuacchiáre, as Viuacchiáre.
Sul, vpon thee, ouer thee. Also in and vpon, in the nicke, at the instant.
Sù leuáre, to rise or raise vp.
Sul far del giórno, vpon breake or by peep of day.
Sul far di nótte, vpon twilight or aproch of night.
Sulfurággine, sulphrinesse.
Sulfurẻo, sulphrie, of sulpher.
Sulimáto, Mercurie or Sublimate that Alchimists vse.
Sullunáre, of or belonging vnder the Moone.
Sultaníno, a kind of Turkish coine.
Sumácchio, Sumach to dye blacke with.
Sumína, a Hogs-panch, or bellie with the paps vnto it, and so drest.
Súmma, as Sómma, a summe.
Summáre, as Sommáre, to summe.
Summário, as Sommário, a summarie.
Summessióne, as Sommessióne.
Summésso, as Somésso.
Summéttere, as Sottométtere.
Summissíuo, as Sommissíuo.
Sumministráre, as Somministráre.
Sumministratióne, as Somministratióne.
Summísta, a Summist, one that followes the opinion of Thomas Aquinas
his summes.
Sumministratóre, as Somministratóre.
Sumptióne, a taking, a receiuing.
Sun, for Se uno, if one. Also for Sù.
Sunáre, to gather vp, or together.
Sunnét, a Turkish word, as much to say as circumcision.
Súnto, a summe, or totall rekoning.
Suntuárie léggi, Lat.
Suntuosaménte, sumptuously, gloriously.
Suntuosità, as Sontuosità.
Suntuóso, as Sontuóso, sumptuous.

Súo, his, his owne, hers, her owne.

Súo, vsed of Dánte for Sù.
Suócera, a wiues or husbands mother.
Suócero, a wiues or husbands father.
Súo dánno, at his perill, or danger.
Suogliáre, to bring out of taste or willingnes, to make vnwilling. Also to
haue or giue one his longing.
Suogliataménte, vnwillingly, vnlustily, sadly, as it were against taste.
Suogliáto, vnwilling, vnlustie, tastelesse, sad, pensiue. Also nice or so
altred, that nothing can please. Also had ones will or longing.
Suói, his, his owne, hers, her owne.
Suóla, a sole of a shoe, &c.
Suolacchiáre, as Suolazzáre.
Suoláio, as Soláio, or Soláro.
Suoláre, as Suolazzáre.
Suoláre, to sole as a paire of shoes. Also to seele, or floore, or roofe any
roome.
Suolazzaménto, as Suolázzo.
Suolazzáre, to beate, to flap, to hoouer or flutter with ones wings. Also to
waue in the winde. Also to pur as a Partridge.
Suolázzi, all manner of wauings or streamings, or banerols, namely, in
Ships or head pieces.
Suolázzo, a beating, a flapping, a fluttring or hoouering with ones wings.
Also a wauing in the winde. Also a purring of a Partridge.
Suolgarizzáre, to translate, to explaine, or make vulgar.
Suóle, he or she is wont or accustomed.
Suolére, vóglio, vólsi, volúto, to nill, to be vnwilling, to vunwill.
Suólgere, vólgo, vólsi, vólto, to turne or diuert away or from.
Suolgiménto, a turning or diuerting away or from.
Suólo, a thing that beareth any thing on it, as the earth, the ground, the
hand, the bottome, the foundation or superficies of any thing. Also a
ground-worke, a pauement, a floore, a seeling, a roofe of any roome or
place.
Suolo a suólo, piled in order, or ranged in ranks one vpon another, as
Hearings be.
Suolontáre, to vnwill, as Suogliáre.
Suólsi, it is wont, the custome or fashion is.
Suólta, as Suolgiménto, a crankling, a winding, a turning.
Suoltáre, to tosse, to turne, to tumble, to welter, or wallow vp and downe,
in or about.
Suólto, turned, or diuerted from, or away.
Suoltoláre, as Suoltáre.
Suolúto, nilled, vnwilled, bin vnwilling.
Suonánza, as Sonánza.

Suonáre, as Sonáre, to sound.

Suonáta, any sound of Musike.
Suonatóre, as Sonatóre.
Suóno, what noise or sound soeuer the eare of man can comprehend.
Suóno a mórto, a knell, a tolling, or ringing, or peale for the dead.
Suóno di pálme, the noise of clapping of hands.
Suóra, a Sister. Also a title giuen to Nunnes or religious women. Also a
Nunne.
Suorésco, Sister-like, Nunne-like.
Supeditáneo, as Suppeditáneo.
Supeditáre, as Suppeditáre.
Supeditatióne, as Suppeditatióne.
Supẻllẻticário, as Suppẻllẻticário.
Supẻllẻtile, as Suppẻllẻtile.
Súpẻr, vp, by, vp along, vp through.
Superábile, vanquishable, that may be surpassed or ouercome.
Supẻrabondánza, superaboundance.
Supẻradditióne, an ouer-addition.
Supẻrallegáto, ouer-bound. Also ouer-alleaged.
Supẻraltatióne, an ouer-extolling.
Supẻrangẻlico, more then angelicall.
Supẻrannáto, out-worne with yeares, out-yeared.
Superánza, as Superiorità.
Superáre, to ouercom, to surpasse, to surmount, to ouer-match, to excell.
Supẻrarogáre, to supererogate, to giue more then is required or expected.
Supẻrarogatióne, supererogation, a giuing more then is required.
Superatióne, an ouercomming, a superation, a surmounting, a surpassing.
Supẻrbaménte, proudly, stately.
Supẻrbia, pride, disdaine, surquedrie, arrogancie, haughtinesse of minde.
Supẻrbiloquẻnte, proudly speaking, a proud or disdainefull speaker.
Supẻrbiloquẻnza, proud or arrogant speach, disdainefull speaking.
Supẻrbíne, the Proud-peare.
Supẻrbio, as Supẻrbo.
Supẻrbióso, full of pride, arrogancie, or disdaine.
Supẻrbíre, bísco, bíto, to make or become proud, disdainefull, haughtie, or
arrogant.
Supẻrbo, proud, disdainefull, surquedrous, arrogant, haughtie, or high-
minded, surlie. Also magnificent, sumptuous, stately, noble,
representing maiestie.
Supẻrbótto, sicke of pride, very proud.
Supẻrbúzzo, a proud squall, a squeamish disdainefull man.
Supẻrcelẻste, more then heauenly.
Supẻrcẻssióne, a supersedies, Lat.
Supẻrchiaménte, as Sopẻrchiaménte.
Supẻrchiáre, as Sopẻrchiáre.
Supẻrchiaría, as Sopẻrchiería.

Supẻrchiatóre, as Sopẻrchiatóre.
Supẻrchiatúra, as Sopẻrchiatúra.
Supẻrchio, as Sopẻrchio.
Supẻrchióso, as Sopẻrchióso.
Supẻrcíglio, the vpper eye-lid of a mans eye, the eye-browes, or ridge of
haires aboue the eye-lids. Also the knop, the brow, the top, the ridge, or
highest part of any great hill, rocke, or crag. Also the seueritie, grauitie,
pride, stately looke, or countenance of a man. Also the hanse of a dore,
or mantle of a chimnie, &c.
Supẻrciglióso, that hath big eye-lids, or eye-browes. Also high-looking,
sur-ie, or grim in countenance, disdainefull, or haughtie-minded.
Supẻreminẻnte, supereminent, being, appearing, or aloft aboue others.
Supẻreminẻnza, a prerogatiue, or excellencie aboue others.
Supẻressentiále, superessentiall.
Supẻressentialità, superessentialitie.
Superéuole, as Superábile.
Supẻrficiále, superficiall, not very deepe.
Supẻrficiáre, to peruse over superficially.
Supẻrficiáro, a superficiall man, one that hath no deepe skill. Also of or
belonging to the out-side of a thing. Also he that buildeth a house on
another mans ground, and therefore payeth quitrent.
Supẻrfície, the surface, the superficies or outside of any thing.
Supẻrfluaménte, with superfluitie.
Supẻrfluáre, to ouerflow, to run ouer, to abound, to exceed, to bee
superfluous.
Supẻrfluíre, ísco, íto, as Supẻrfluáre.
Supẻrfluità, superfluitie, more then needes, profusenesse, ouerplus.
Supẻrfluo, superfluous, ouer-much, enough, and much to spare, profuse,
that runneth ouer.
Supẻrfúgio, a flying away.
Supẻrfusióne, as Embrocatióne.
Supẻrfúso, powred vpon, ouerflowed.
Súperi, as Superióri.
Supẻrimpendẻnte, hanging ouer.
Supẻrinfóndere, to infuse from aboue.
Supẻrinfúso, infused from aboue.
Superióre, superior, vppermost, former, senior, higher, chiefer, elder, in
better place.
Superióri, superiors, seniors, elders, ancestors, predecessors. Also Gods,
Angels, Saints, or they that are in Heauen.
Superiorità, superioritie, preheminence, senioritie, abouenesse, eldership.
Supẻrlatióne, the figure hiperbole, or excessiue amplifying. Also a surname
giuen to one by excellencie.
Supẻrlatiuaménte, superlatiuely, in the highest degree.

Supẻrlatíuo, a superlatiue, or chiefe degree.

Supẻrláto, excessiuely amplifide. Also preferred, borne vpon or ouer.
Supẻrlimináre, the lintle ouer a dore.
Supẻrnále, supernall, high, aboue, chiefe.
Supernalménte, supernally.
Supẻrnatie, a kind of rare Peaches.
Supẻrnáto, borne or growne ouer, or aboue.
Supẻrnaturále, supernaturall.
Supẻrnaturalità, supernaturalitie.
Supẻrno, supernall, aboue, highest.
Súpero, as Superióre.
Supẻrsáta, a kind of Haggas, or Pudding.
Supẻrsedére, as Soprasedére.
Supẻrsẻdie, as Soprasẻdie.
Supẻrspicáce ingégno, a superspicuous wit.
Supẻrste, as Supẻrstíto.
Supẻrstitióne, superstition, curiositie, scruple, a vaine religion or deuotion,
an honouring of that should not be honoured, a spiced conscience in
vaine things.
Supẻrstitióso, superstitious, scrupulous, curious, vaine or timorous in
conscience, or religion without cause.
Supẻrstito, escaped and safe from death, a suruiuor, one that ouerliueth
others.
Supẻruacáno, out-worne with yeeres, out-yeered.
Supẻruácuo, vaine, to no vse, vnnecessarie, superfluous or ouermuch.
Supináre, to lay vpright or the bellie vpward. Also to goe stately or proudly
as with the nose in the winde.
Supináto, laid vpright, turned vp, holden vp in the ayre. Also somewhat
bowed backward.
Supinità, lying vpright with the face vpward. Vsed also for negligence.
Supíno, vpright, with the face or belly vpward. Also idle, lither or
retchlesse. Also a Supine deriued of a verbe.
Supirióso, as Suspirióso, sighfull.
Suplantáre, to supplant.
Suplantatióne, a supplantation.
Suplantatóre, a supplanter.
Súplica, as Súpplica, &c.
Suplicáre, as Supplicáre.
Supositório, a suppository.
Súppa, as Sóppa, any kind of sop.
Suppáre, as Soppáre, to sop.
Suppedáneo, any thing set vnder the feet, a foot-stoole, a thing that
images stand vpon.
Suppeditáre, to haue sufficient, to be able to giue and furnish to others.
Also to subdue or bring low to ones feete.
Suppeditatióne, a giuing or ministring vnto another what he lacketh.

Suppelẻticário, a slaue or bondman which is accounted for chattle. Also a

seruant hauing a charge of houshold stuffe.
Suppelẻtile, all things mooueable in a house, all manner of houshould
stuffe, goods or chattle or implements.
Suppiláre, to steale, to pilfer or purloine vnder hand, to rob or filch craftily.
Suppilatióne, a stealing or pilfering.
Supplantáre, to supplant.
Supplantatióne, a supplantation.
Supplantatóre, a supplanter.
Súpplica, a supplication, a sute, a request, an entreatie. Also a generall
procession.
Supplicánte, humblie intreating or beseeching. Also a suppliant, or sutor.
Supplicáre, humblie to sue, to beseech or intreate by way of supplication.
Supplicatióne, as Súpplica.
Supplicatória lẻttera, a letter of beseeching, or earnest suing.
Súpplice, humblie intreating, suing, or beseeching, suppliantly, prostrately
crauing.
Suppliceménte, in humble kneeling or beseeching manner, suppliantly
crauing.
Supplichéuole, as Súpplice.
Supplicheuolménte, as Suppliceménte.
Supplício, punishment, torture or torment giuen or suffred for any offence.
Supplicióso, full of torture or punishment.
Suppliménto, a supplie of any want, a filling vp the number and places.
Supplíre, plísco, plíto, to supply or furnish any want, to make vp that
which lacketh, to fill vp the number or place of what lacketh.
Supplíta, as Suppliménto.
Suppónere, to suppose or surmise. Also to put or set vnder.
Suppórre, as Suppónere.
Suppositalménte, supposedly.
Suppositítie, matters vpon surmise.
Suppositióne, a supposition. Also a puting of one thing vnder another.
Suppositítio, counterfeit, supposed. Also that may be supposed.
Supositíuo, suppositiue, that may be supposed or surmised.
Suppósito, vnderset or put vnder. Also a suppose or thing supposed.
Suppositório, a suppository.
Suppósto, as Suppósito. Also supposed.
Suppuránza, as Suppuratióne.
Suppuráre, to grow to an impostume or matter as a sore doth.
Suppuratióne, a gathering to rotten matter, an impostumation.

Suppuratíuo, a salue to rot a sore or to bring it to run matter.

Supputáre, to suppose, to deeme or surmise. Also to count, to compute,
to reckon. Also to prune or cut trees slightly.
Supputatióne, a supposing or surmise. Also a computation or reckoning.
Also a pruning of trees.
Súpra, hath beene vsed for Sópra.
Supremità, supremity, supremacy, superiority, soueraignety, chiefe-dome.
Suprẻmo, supreme, chiefe, soueraigne, highest or greatest of all. Also a
mans last will and testament. Also death or funerall obsequies or latter
daies.
Súr, as Sópra, ouer, namely vsed in words composed, as Surmontáre, &c.
Súra, a kind of precious liquor that distilleth from the branches that are
riuen or cut from that tree that beareth the Indian great white nut,
which the Indians vse to drinke, being new it is called Súra, but being
setled and stall it is called Otráqua. Also the whole shanke or legge from
the knee to the anckle. Also the calfe of the leg. Some haue also vsed
Súra for the lesser bone of a mans legge or shanke.
Súrco, hath been vsed for Sólco.
Surculário, a kinde of greene Grasse-hopper.
Súrculo, a yongue slip, set, graffe, scien, shoote, twig, branch or burgeon.
Súrdo, as Absúrdo, absurd. Also as Sórdo.
Surẻlla, a fish called in Latine Lacertus.
Surétti, high corked chopinos or pantofles.
Surẻttióne, as Resurẻttióne.
Surẻtto, risen vp, as Resurẻtto.
Surétto, a kinde of compounded drinke vsed in Turkie instead of wine.
Surgẻnte, rising, springing or growing vp. Also riding at anchor as a ship.
Súrgere, súrgo, súrsi, súrto, as Sórgere.
Súrgere in pórto, a ship to rid in an hauen or to come to anchor in any
Port.
Surgitóre, as Sorgitóre, as Surgitório.
Surgitório, an anchoring place, a bay or rode for ships to ride in or at.
Surgíuo, as Sorgíuo, as Sortíuo.
Súrgo-Túrco, a kinde of cource graine or Corne.
Surlunáre, whatsoeuer is aboue the Moone.
Surmáre, as Ciurmáre, to hale, to halse, to tug.
Surmontáre, to surmount.
Surmontéuole, surmountable.
Súrnag, a kind of medicinable roote.
Súro, as Surẻlla.
Surprẻndere, to surprise or take on a suddaine.

Surprésa, a surprise or taking on a suddaine.

Surrẻntíno, a kind of wine.
Surrẻttióne, a rising vp.
Surrẻtto, risen vp.
Súrro, the wood Corke.
Surrogáre, to subrogate, to substitute, to make a Deputie, to appoint in
anothers roome.
Surrogatióne, a subrogation, a substituting or appointing in place of
another.
Súrse, he did rise vp, he arose. Also vsed of Dánte for Súso.
Súrso, as Sórso. Also vp or vpward.
Súrto, risen, growen or sprung vp. Also ridden at anchor or in a rode as
ships doe.
Suruissúto, suruiued, ouer-liued.
Suruíuere, to suruiue, to ouerliue.
Suruoláre, to ouer-flie.
Suruólo, an ouer-flight.
Susaiuólo, a wherne for a spindle.
Susaméli, a kind of daintie Plum.
Suscẻttíbile, that may be enterprised or vndertaken.
Suscẻttióne, an enterprise or vndertaking.
Suscẻtto, enterprised or vndertaken. Also begotten. Also receiued.
Suscíno, as Susíno.
Suscípere, to beget, to conceiue. Also to vsurpe, to vndertake or
enterprise.
Suscitábolo, a prouocation or stirring vp.
Suscitaménto, as Suscitatióne.
Suscitáre, to raise or rouze vp as from out a sleepe. Also to exhort, to
prouoke or stirre vp. Also to mooue or stirre vp vnto some insurrection.
Suscitatióne, a raising or rouzing vp. Also an exhorting or prouoking vnto.
Also an vprore or insurrection.
Suscitatóre, a raiser, a stirrer vp. Also a seditious man, a prouoker.
Susimáni, a kind of Plum in India.
Susína, a Damzen-plum or prune.
Susináro, a Plum or Damzen-tree.
Susíne sécche, dried Prunes.
Susíno, a Plum or Damzen-tree. Also an oyle of Lillies.
Susíno seluático, a Skeg-tree.
Súso, vpon, ouer on or aboue. Also an Aduerbe of encouraging, as we say,
vp, on, goe to, away, forward, lustily.
Susornáre, to thump or whirret on the head with the fist.
Susórno, a thump or whirret giuen with the fist vpon ones head.
Suspẻndere, as Sospẻndere.
Suspẻndio, a hanging vp or by.
Suspẻnsióne, as Sospẻnsióne.
Suspéso, as Sospéso.
Suspẻttáre, as Sospẻttáre.
Suspẻtto, as Sospẻtto.

Suspẻttóso, as Sospẻttóso.
Suspicáre, as Sospẻttáre.
Suspicóso, as Sospẻttóso.
Suspírio, as Sospírio.
Suspirióso, sighfull, panting, purcie, short-winded.
Suspittióne, as Sospittióne.
Susseguẻnte, subsequent, next following or succeeding.
Susseguẻnteménte, succeedingly.
Susseguẻnza, a subsequence, a consequence.
Sussequẻnte, as Susseguẻnte.
Sussidiáli, as Sussidiárij.
Sussidiáre, to helpe, to aide, to relieue, to supply, to stand by to aide and
defend. Also to subsidie or taske with subsidies.
Sussidiárij soldáti, subsidiarie souldiers or such as serue for defence of
others, or that come or are sent to aide others. Also of or belonging to
subsidies or helpes.
Sussídio, subsidie, helpe, reliefe, supply, assistance. Also a refuge or
rescue. Also aide or succour at a pinch or need.
Sussiẻgo, as Sossiẻgo.
Sussistẻnte, subsisting.
Sussistẻnza, subsistence.
Sussístere, sísto, sistéi, sistúto, to subsist.
Sussoláre, of or belonging aboue the Sunne.
Sussuráre, to whisper, to mutter or murmure. Also to humme or buzze as
Bees. Also to charme or forespeake with whispring words. Also to make
a low, a soft or still noise as a gentle winde among trees, or a gentle-
gliding streame among pible-stones, or as birds chirping and chattring
among woods.
Sussuratióne, a priuie whispring, murmuring, humming or buzzing.
Sussuratóre, a priuie whisprer, murmurer, buzzer, hummer, charmer or
muttrer, a secret carie-tales, a priuie backbiter, &c.
Sussúrro, a priuie whispring, a soft murmure, a low humming or buzzing,
a soft still noise of a gentle winde, of a running streame, of chirping
birds.
Sussurróne, as Sussuratóre.
Sussurroneggiáre, as Sussuráre.
Sústa, the name of one of the chiefe cables in a ship.
Sustánte, vp-standing, standing vpright.
Sustántia, as Sostánza.
Sustantiále, as Sostantiále.
Sustantíuo, as Sostantiuo.
Sustáre, to binde vp the sailes of a Ship.
Susteniménto, as Sosteniménto.
Sustentácolo, that holdeth vp.
Sustentáre, as Sostenére.

Sustentaménto, as Sosteniménto.
Sustentatióne, as Sostentatióne.
Susterẻllo, ci restórono i dẻnti susterẻlli.
Sustituíre, as Sostituíre.
Sustituíto, as Sostituíto.
Sustitutióne, as Sostitutióne.
Sustitutíuo, that may be substituted.
Sustitúto, as Sostituíto.
Susuráre, as Sussuráre, &c.
Susúrno, as Olfattório.
Sutéla, subtletie, craftie working, wilie deceite, cauillation inuented with
purpose to deceiue a couenant.
Sutelóso, a craftie wrangler, a subtle cauiller, one full of cauillations.
Súto, as Státo, the Participle of Éssere, bin. Also sewed or seamed.
Sutóre, a Shoemaker, a Souter.
Sutório, souterlie, shoemaker-like.
Suttẻrfuggíre, as Sottẻrfuggíre.
Suttẻrfúgio, as Sottẻrfúgio.
Suttilità, as Sottilità.
Suttilizzáre, as Sottigliáre.
Sútto, dry, dried, wiped dry. Also suckt or drawen dry. Also barren.
Sutúra, a suture, a seame, a fastning or ioyning together. Also the twist or
line vnder a mans yard. Also a ioyning together of bones in the head,
like to a thing drawen together with long stitches: or that manner of
composition seene in the head, wherein the bones being indented one
within another, are like the teeth of a Sawe.
Sutúra rẻtta. Looke Sagittále.
Súuaro, as Súuero.
Suuerétto, a place full of Corke-trees.
Súuero, the tree or wood Corke.
Súui, for Iui sópra, thereupon.
Sù uía, on away, vp and away.
Suuotáre, to emptie.

T.

Tà, as Túa, thine, thine owne.

Ta', for Táli, such, such like.
Tabaccáre, to take Tabacco. Also as Tabacchináre.
Tabacchináre, to play the secret pander or close bawde namely about
boies.
Tabacchíno, a secret pander for boies.
Tabácco, the hearbe Tabacco. Also a bunch or bumpe of flesh growing
vnder the skin.
Tabaliáro, a Drummer, a Tabourer.
Tabálli, Drums, Drumslades, Tabours or Timbrels.
Tabanẻllo, as Gabanẻllo.
Tabáno, as Taffáno.
Tabáro, as Tabárro.
Tabarráre, to cloake or couer with a riding cloake that hath a hood.
Tabarríno, a short riding cloake with a hood. Also a Shepheards frocke.
Tabárro, a riding cloake or Shepheards frocke with a hood.
Tábe, a pining, a withering, a consuming or wasting of the body by long
sicknesse, or for want of nourishing and lacke of naturall moistnesse, a
tissike, a consumption and putrifying of the lungs. Also infection. Also
putrefaction, rotten matter or corrupt bloud comming out of a wound or
cankered sore. Vsed also for strength to putrifie and consume.
Tabefáre, to pine away, to corrupt, to consume or putrifie.
Tabefattióne, as Tábe.
Tabẻlla, a writing table or tables.
Tabẻllionáto, Scriuener-ship.
Tabẻllióne, a Scriuener or publike Notarie.
Tabẻnto, a kind of spungious stone.
Tabẻrciáre, to patch, to botch, to piece, to clout, to mend.
Tabẻrciatíua árte, the arte of patching, botching, clouting, pieceing or
souting.
Tabẻrciatóre, a patcher, a botcher, a clouter, a piecer, a tinkar, &c.
Tabẻrnácolo, a Tabernacle, a shrine, a cubbord or box or chariot to carry
or keepe images in. Also a Tent or a Pauilion or shed made of boords.
Tabẻrnaríe, Comedies plaid by common base people, pertayning to
Tauernes.
Tabẻrníre, nísco, níto, as Tabẻscere.
Tabẻscere, bẻsco, bẻscéi, bẻscíuto, to consume, to pine and waste away
as by a consumption. Looke Tábe.
Tabẻsciúto, consumed, wasted or pined away. Also corrupted and rotten.
Tabetúra, as Tábe.
Tábi, the stuffe called Taby or Tabine.
Tábido, as Tabesciúto.
Tabífico, that bringeth into or causeth a consumption. Looke Tábe.
Tabíno, a stuffe called Tabine.
Tabíto, as Tabẻsciúto.
Tabitúdine, as Tábe.
Tábo, corrupt, goare, or rotten blood.
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookbell.com