1.

Introduction to Machine
Learning
Subject : Machine Learning(3170724)
Faculty: Dr. Ami Tusharkant Choksi
Associate professor, Computer Engineering Department,
Navyug Vidyabhavan Trust
C.K.Pithawala College of Engineering and Technology,
Surat, Gujarat State, India.
Website: www.ckpcet.ac.in

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724)

 Books
1. Machine Learning, Saikat Dull, S. Chjandramouli, Das, Pearson
2. Machine Learning with Python for Everyone, Mark Fenner, Pearson
3. Machine Learning, Anuradha Srinivasaraghavan, Vincy Joseph, Wiley
4. Machine Learning with Python, U Dinesh Kumar Manaranjan Pradhan, Wiley
5. Python Machine Learning, Sebastian Raschka, Vahid Mirjalili, Packt Publishing

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 2

 Contents
- Overview of Human Learning and Machine Learning
- Types of Machine Learning
- Applications of Machine Learning
- Tools and Technology for Machine Learning

02 hours

CO1: Explore the fundamental issues and challenges in Machine Learning

including data and model selection and complexity.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 3

 Human learning
● Learning happens when one observe a phenomena and recognize a pattern. One
try to understand this pattern by finding out if there is any relationship between
the entities involved in that phenomena.
● Human learning is a process of acquiring knowledge. Our behavior, skills, values
and ethics are acquired when we process information through our minds and
learn. Human learning may occur as part of education, personal development or
any other informal/formal training.
● Learning is the process of acquiring new understanding, knowledge, behaviors,
skills, values, attitudes, and preferences.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 4

 Human learning example color=white, weight=800 kg, trunk, tail,
ears shape, eyes, teeth

color=black
weight=800 kg,
trunk,
color=black, weight=6800 kg, trunk, tail, tail
ears shape, eyes, teeth
Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 5
 Machine learning
● Machine learning (ML) is the study of computer algorithms that improve
automatically through experience and by the use of data.
● It is seen as a part of artificial intelligence.
● Machine learning algorithms build a model based on sample data, known as
"training data", in order to make predictions or decisions without being
explicitly programmed to do so.
● Machine learning algorithms are used in a wide variety of applications, such as
in medicine, email filtering, speech recognition, and computer vision, where it is
difficult or unfeasible to develop conventional algorithms to perform the needed
tasks.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 6

 Machine learning definition
● “Learning is any process by which a system improves performance from
experience”.-Herbert Simon in 1975
● “Machine learning is concerned with machines that automatically improves their
performance through experience.”-Herbert Simon in 1975
● Machine learning is a science of getting computers to act without being explicitly
programmed.-stanford university
● Machine learning algorithms can figure out how to perform important tasks by
generalizing from examples.
● Machine learning seeks to answer the question, “how to build a system that
automatically improves with experience and fundamental laws that govern all
the process?”

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 7

 Human vs. machine learning
Feature Human learning Machine learning

Learning Humans acquire knowledge Machines acquire knowledge through experience

through experience either shared in the form of past data.
directly or shared by others.
Typically knowledge, Skill, and
Memory being used to define
intelligence.

Intelligence Humans without knowledge in Machines can solve new problems only if their
particular subjects can apply intelligence has been updated with retraining on
their intelligence to solve data acquired from the changed scenarios.
problems in new domains.

Mistake Both humans and machines make mistakes in applying their intelligence in solving
(similarity) problems. In ML, overfitting memorizes all examples and an overfitted model lacks
generalization and it fails to work on never seen before examples
Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 8
 Human vs. machine learning
Feature Human learning Machine learning

Transfer In humans, transfer of In ML, Transfer Learning is a technique that

learning knowledge to students is reuses a model that was created by machine
typically by teachers. learning experts and that has already been
trained on a large dataset.

Making Transfer learning may not make Transfer learning makes the transferee as
intelligent the students intelligent. It intelligent as the transferor. Transfer learning
depends on the inherent leverages information extracted from one set of
intelligence of the transferee to distributions.
enhance his/her problem
solving skills.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 9

 Machine learning is an intersection

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 10

 Artificial Intelligence
● Artificial intelligence (AI) is intelligence demonstrated by
machines, unlike the natural intelligence displayed by humans
and animals, which involves consciousness and emotionality.
● The distinction between the former and the latter categories is
often revealed by the acronym chosen. 'Strong' AI is usually
labelled as artificial general intelligence (AGI) while attempts to
emulate 'natural' intelligence have been called artificial
biological intelligence (ABI).
● Artificial intelligence (AI) is the ability of a computer or a robot
controlled by a computer to do tasks that are usually done by
humans because they require human intelligence and
discernment(સમજદારી).

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 11

 Machine learning and data mining
● Machine learning and data mining often employ the same methods and overlap
significantly, but while machine learning focuses on prediction, based on known
properties learned from the training data, data mining focuses on the discovery
of (previously) unknown properties in the data (this is the analysis step of
knowledge discovery in databases).
● Data mining uses many machine learning methods, but with different goals; on
the other hand, machine learning also employs data mining methods as
"unsupervised learning" or as a preprocessing step to improve learner accuracy.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 12

 Types of Machine Learning
1. Supervised learning
2. Unsupervised learning
3. Semi supervised learning
4. Reinforcement learning

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 13

 Supervised vs. Unsupervised learning

Image src: https://static.javatpoint.com/tutorial/machine-

learning/images/difference-between-supervised-and-unsupervised-
learning.jpg
Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 14
 Supervised vs. Unsupervised learning

Feature Supervised learning Unsupervised learning

Labeled/Unla Algorithms are trained using Algorithms are trained using

beled Data labeled data. unlabeled data.

Data Input data is provided to the Only input data is provided to

with/without model along with the output. the model.
Output

Groups Classes are already defined Groups are made based on

in data using labels similar characteristics.

Feedback of Supervised learning model Unsupervised learning model

output takes direct feedback to does not take any feedback.
check if it is predicting
correct output or not.
Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 15
 Supervised vs. Unsupervised learning

Feature Supervised learning Unsupervised learning

Predicts Supervised learning model Unsupervised learning model

output/pattern predicts the output. finds the hidden patterns in data.

Goal The goal of supervised The goal of unsupervised learning

learning is to train the model is to find the hidden patterns and
so that it can predict the useful insights from the unknown
output when it is given new dataset.
data.

Accuracy of Supervised learning model Unsupervised learning model may

results produces an accurate result. give less accurate result as
compared to supervised learning.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 16

 Supervised vs. Unsupervised learning

Feature Supervised learning Unsupervised learning

Algorithms Linear Regression, Logistic It includes various algorithms

Regression, Support Vector such as Clustering, KNN, and
Machine, Multi-class Apriori algorithm.
Classification, Decision tree,
Bayesian Logic, etc.

● Clustering is of two types: hard and soft.

○ Hard clustering is where each data point is part of only one cluster
○ Soft clustering is where each data point is part of more than one
cluster.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 17

 Clusters

Hard Clustering Soft Clustering

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 18
 Semi supervised learning
● Semi-supervised learning is an approach to machine learning that combines a
small amount of labeled data with a large amount of unlabeled data during
training. Semi-supervised learning falls between unsupervised learning (with no
labeled training data) and supervised learning (with only labeled training data).

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 19

 Reinforcement learning
● Reinforcement learning is an area of machine learning concerned with how
intelligent agents ought to take actions in an environment in order to maximize
the notion of cumulative reward.
● Example: The problem is as follows: We have an agent and a reward, with many
hurdles in between. The agent is supposed to find the best possible path to reach
the reward.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 20

 Reinforcement learning: Positive
● Types of Reinforcement: There are two types of Reinforcement:
● Positive Reinforcement is defined as when an event, occurs due to a particular
behavior, increases the strength and the frequency of the behavior. In other
words, it has a positive effect on behavior.
● Ladder - benefit-positive
● Advantages of reinforcement learning are:
○ Maximizes Performance
○ Sustain Change for a long period of time
● Disadvantages of reinforcement learning:
○ Too much Reinforcement can lead to overload of states which can diminish the results

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 21

 Reinforcement learning: Negative
● Negative Reinforcement is defined as strengthening of a behavior because a
negative condition is stopped or avoided.
● Advantages of reinforcement learning:
○ Increases Behavior
○ Provide defiance to minimum standard of performance
● Disadvantages of reinforcement learning:
○ It Only provides enough to meet up the minimum behavior

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 22

 Example of reinforcement learning
● When ladder is there +ve points are
there - award, reward
● When snake is there -ve points are
there-penalty

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 23

 Applications of Machine Learning

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 24

 Tools and Technology for Machine Learning
● Knime
○ open-source GUI based machine learning tool
○ doesn’t require any knowledge of programming
○ used for data manipulation, data mining
● Accord.net:
○ computational machine learning framework
○ an image as well as audio packages
○ this framework is C# language

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 25

 Tools and Technology for Machine Learning
● Scikit-Learn:
○ an open-source machine learning package.
○ unified platform as it is used for multiple purposes.
○ assists in regression, clustering, classification, dimensionality reduction,
and preprocessing.
○ built on top of the three main Python libraries viz. NumPy, Matplotlib,
and SciPy. Along with this, it will also help you with testing as well as
training your models.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 26

 Tools and Technology for Machine Learning
● TensorFlow:
○ an open-source framework that comes in handy for large-scale as well
as numerical ML.
○ It is a blender of machine learning as well as neural network models.
○ it is also a good friend of Python.
○ it runs on CPU and GPU as well.
○ Natural language processing, Image classification are the ones who
implement this tool.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 27

 Tools and Technology for Machine Learning
● Weka:
○open-source software.
○Has a graphical user interface.
○is very user-friendly.
○The application of this tool is in research and teaching.
○Along with this, Weka lets you access other machine learning tools as
well. For example, R, Scikit-learn, etc.
● Pytorch :
○ a deep learning framework.
○ very fast as well as flexible to use, because it has a good command over
the GPU.
Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 28
 Tools and Technology for Machine Learning
● Pytorch :
○ it is used in the most vital aspects of ML which includes building deep
neural networks and tensor calculations.
○ is completely based on Python.
○ Along with this, it is the best alternative to NumPy.
● RapidMiner :
○ a deep learning framework.
○ platform-independent
○ their own data as well as test their own models
○ drag and drop working

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 29

 Tools and Technology for Machine Learning
● Google Cloud AutoML:
○ The objective of Google cloud AutoML is to make artificial intelligence
accessible to everyone.
○ it provides the models which are pre-trained to the users in order to
create various services. For example, text recognition, speech
recognition, etc.
○ popular among companies, as the companies want to apply artificial
intelligence in every sector of the industry but they have been facing
difficulties in doing so because there is a lack of skilled AI persons in the
market.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 30

 Tools and Technology for Machine Learning
● Jupyter Notebook :
○ most widely used machine learning tools
○ very fast processing as well as an efficient platform.
○ it supports three languages viz. Julia, R, Python.
○ Thus the name of Jupyter is formed by the combination of these three
programming languages.
○ allows the user to store and share the live code in the form of
notebooks.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 31

 Tools and Technology for Machine Learning
● Apache Mahout:
○ is launched by Apache which is an open-source platform based on
Hadoop.
○ used for machine learning and data mining.
○ techniques such as regression, classification, and clustering became
possible
○ it also makes use of math-based functions such as vectors, etc.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 32

 Tools and Technology for Machine Learning
● Azure Machine Learning studio:
○ is launched by Microsoft.
○ provides machine learning services to the users.
○ is a very easy way to form connections of modules and datasets.
○ provides AI facilities to the user. Just like TensorFlow, it also works on
CPU and GPU.
● MLLIB:
○ Like Mahout, MLLIB is also a product of Apache Spark.
○ used for regression, feature extraction, classification, filtering, etc.
○ called Spark MLLIB.
○ very good speed as well as efficiency.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 33

 Tools and Technology for Machine Learning
● Orange3:
○ a data mining software
○ assists in preprocessing, data visualization, and other data-related stuff.
○ can access through the Anaconda Navigator.
○ very helpful in Python programming.
○ a great user interface.
● IBM Watson :
○ a web interface is given by IBM for using Watson.
○ Watson is a human interaction Q and A system which is based on
Natural Language processing.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 34

 Tools and Technology for Machine Learning
● IBM Watson :
○ is applied in various fields such as automated learning, information
extraction, etc.
○ is used for research and testing purposes.
○ Its objective is to offer a human-like experience to the users.
● Pylearn2:
○ is a machine learning library that is built on top of Theano.
○ Therefore, there are many functions that are similar between them.
○ it can perform math calculations.
○ Is capable of running on the CPU and GPU as well. Before getting to
Pylearn2, you must be familiar with Theano.

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 35

 Issues and Challenges in machine learning

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 36

 Practical Issues in machine learning
● Lack Of Quality Data: noisy, incorrect, incomplete data
● GETTING BAD RECOMMENDATIONS : proposal engines with complex algorithms
may tend to provide wrong results
● TALENT DEFICIT:rare experts are available
● IMPLEMENTATION: slow deployment, data security, lack of data
● MAKING THE WRONG ASSUMPTIONS
● DEFICIENT INFRASTRUCTURE
● HAVING ALGORITHMS BECOME OBSOLETE WHEN DATA GROWS
● ABSENCE OF SKILLED RESOURCES
● CUSTOMER SEGMENTATION
● COMPLEXITY
● LACK OF EXPLAINABILITY

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 37

 Challenges in machine learning
● Data Collection
● Less Amount of Training Data
● Non-representative Training Data
● Poor Quality of Data
● Irrelevant/Unwanted Features
● Overfitting the Training Data
● Underfitting the Training data
● Offline Learning & Deployment of the model

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 38

 References
1. Introduction to machine learning, https://www.slideshare.net/liorrokach/introduction-to-machine-
learning-13809045
2. Human learning vs. machine learning, https://towardsdatascience.com/human-learning-vs-machine-
learning-dfa8fe421560
3. Learning, https://en.wikipedia.org/wiki/Learning
4. Machine learning, https://en.wikipedia.org/wiki/Machine_learning
5. Human vs. Machine learning, https://en.wikipedia.org/wiki/Artificial_intelligence
6. Human vs. Machine learning, https://www.datasciencecentral.com/profiles/blogs/human-learning-
and-machine-learning-how-they-differ
7. Types of learning, https://www.ecloudvalley.com/machine-learning/
8. Machine learning tools and technology, https://dzone.com/articles/15-most-used-machine-learning-
tools-by-experts
9. Applications of machine learning, https://www.javatpoint.com/applications-of-machine-learning
10. Supervised vs. Unsupervised learning, https://www.javatpoint.com/difference-between-supervised-
and-unsupervised-learning

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 39

 References
11. Semi supervised machine learning,
12. Challenges in machine learning, https://towardsdatascience.com/top-8-challenges-for-machine-
learning-practitioners-c4c0130701a1
13. Issues in machine learning, https://www.jigsawacademy.com/blogs/ai-ml/issues-in-machine-learning

Dr. Ami T. Choksi @CKPCET Machine Learning (3170724) 40