COURSE CODE COURSE TITLE L T P C

10211CS223 MACHINE LEARNING TECHNIQUES 2 0 2 3

A. Preamble
This course introduces students to new and actively evolving interdisciplinary field of modern data
analysis. The interplay between known data and unknown ones give rise to complex pattern structures
and machine learning methods that are the focus of the study. In the course we will consider methods
of Machine Learning and Data Mining Learning. It also includes dimensionality reduction techniques
which help to deal with huge datasets.

B. Prerequisite Course
10210MA103- Probability, Statistics and Queuing theory

C. Course Objectives
Learners are exposed to:
Apply the concepts of supervised and unsupervised learning algorithms for real time
applications
Experiment classification and association models to overcome the problem of over fitting
Analyse and suggest appropriate machine learning approaches for various types of problems
Demonstrate the aspects of simple perceptron and multi-layer perceptron.

D. Course Outcomes
Upon the successful completion of the course, students will be able to:
CO
Course Outcomes K - Level

Examine the basic concepts of data mining and machine learning

CO1 K2
concepts
Design and evaluate the dimensionality reduction algorithms using real
CO2 K3
world datasets.
CO3 Apply various algorithms of Classification and Association. K3
Demonstrate experiments to evaluate and compare different
CO4 K3
unsupervised learning algorithms
Use the concept of neural networks for learning linear and non-linear
CO5 K3
activation functions

K1-Remember K2-Understand K3-Apply K4-Analyze K5-Evaluate K6-Create

E. Correlation of Cos with Program Outcomes and Programme Specific Outcomes:

Cos P PO PO PO PO PO PO PO PO PO1 PO1 PO12 PSO1 PSO PSO

O 2 3 4 5 6 7 8 9 0 1 2 3
1
CO1 2 2 1 1 3
CO2 3 3 3 2 2 1 1 3
CO3 3 3 3 2 2 1 1 3 2
CO4 3 3 3 2 2 1 1 3 2
93
CO5 3 2 2 2 2 1 1 1 1 3 2
3- High; 2-Medium; 1-Low

F. Course Contents

Unit 1 Introduction to Data mining and Machine Learning L-6 Hours + 3 T Hours
Introduction to Data mining and Knowledge Discovery Data Data Mining Functionalities Steps
in Data Mining Process, Architecture of Typical Data Mining systems - Interesting Pattern -
Classification of Data Mining Systems Machine Learning -Types of machine learning Model
selection and generalization- Evaluation Metric-Perspectives and Issues in Machine Learning
Concept Learning Task Applications of Data Mining and Machine Learning

Unit 2 Dimensionality Reduction L-6 Hours + 3 T Hours

Data Pre-processing- Needs Pre-processing the Data- Data Cleaning, Data Integration and
Transformation, Data Reduction, Discretization and Concept Hierarchy Generation- Dimensionality
Reduction Feature Extraction- Variable Selection- Variable ranking- Linear Discriminant Analysis
Principal Component Analysis Factor Analysis Cross Validation Resampling methods

Unit 3 Association and Classification L-6 Hours + 3 T Hours

Association Rule Mining- Mining Multilevel Association Rules- Mining Multidimensional
Association Rule-Constraint Based Association Mining Decision Tree Induction - Bayesian
Classification Support Vector Machines Linear Regression - Logistic Regression- Decision Trees
Univariate Tree Multivariate Tree Regression Trees ID3- Random forest- Probability and
Learning Gaussian Mixture Models Nearest Neighbour Methods

Unit 4 Clustering Techniques L-6 Hours + 3 T Hours

Cluster Analysis - Types of Clustering methods Partitioning Clustering- K-means, K- Medoid, K
Mode- Density-Based Clustering-DBSCAN Algorithm - Distribution Model Based Clustering -
Maximum likelihood Estimation- Expectation-Maximization Algorithm- Hierarchical Clustering-
Agglomerative Hierarchical algorithm- Fuzzy Clustering- Fuzzy C-means algorithm, Agglomerative
and Divisive Clustering BRICH - ROCK- Chameleon

Unit 5 Neural Networks L-6 Hours + 3 T Hours

ANN- Simple Perceptron- Gate implementation- Multi-layer Perceptron Going Forwards Going
Backwards: Back Propagation Error Multi-layer Perceptron in Practice Examples of using the MLP
Overview Deriving Back-Propagation- Optimization- Hyper parameter tuning

Total:45 Hours

Laboratory Experiments
Part-1

TASK 1 Apply and demonstrate the FIND-S algorithm for finding the most specific Hypothesis
based on a given set of training data samples.
Tools: Google co-lab, Python, Scikitlearn, Anaconda navigator

TASK 2 Implement and demonstrate the Candidate-Elimination algorithm to output a

description of the set of all hypotheses consistent with the training examples. Tools:
Google co-lab, Python, Scikitlearn, Anaconda navigator

94
TASK 3 Apply LDA algorithm to select the appropriate data from the given data set. Use XG
boost algorithm for classification.
Tools: Google co-lab, Python, Scikitlearn, Anaconda navigator

TASK 4 Build a classification model that can effectively analyze and extract features from an
image. Apply PCA algorithm to find the appropriate feature.
Tools: Google co-lab, Python, Scikitlearn, Anaconda navigator

TASK 5 Write a program to demonstrate the working of the decision tree based on ID3
algorithm. Use an appropriate data set for building the decision tree and apply this
knowledge to classify a new sample.
Tools: Weka, Python, Scikitlearn, Anaconda navigator

TASK 6 Write a program to implement the Gaussian mixture model. Compute the accuracy of
the classifier, considering few test data sets.
Tools: Rapid Miner, Python, Scikitlearn, Anaconda navigator
TASK 7 Apply unsupervised learning to implement K means clustering to perform cluster
analysis, find the optimal number of clusters, identify appropriate features and interpret
results.
Tools: Weka, Python, Scikitlearn, Anaconda navigator
TASK 8 Apply unsupervised learning to implement Hierarchical clustering to perform cluster
analysis, find the optimal number of clusters, identify appropriate features and interpret
results.
Tools: Google co-lab, Python, Scikitlearn, Anaconda navigator
TASK 9 Apply EM algorithm to cluster a set of data stored in a .CSV file. Use the same data set
for clustering using k-Means algorithm.Compare the results of these two algorithms
and comment on the quality of clustering.
Tools: Google co-lab, Python, Scikitlearn, Anaconda navigator
TASK 10 Apply back propagation neural network on image data. The idea is to build a Artificial
Neural Network model that can effectively analyze and extract features from an image.
Tools: Google co-lab, Python, Scikitlearn, Anaconda navigator
Task 11 Assuming a set of documents that need to be classified, use the Radial basis function
Classifier model to perform this task. Calculate the classification rate, accuracy,
precision, and recall for your data set.
Tools: Google co-lab, Python, Scikitlearn, Anaconda navigator
Task 12 Write a program to implement artificial neural network with back propagation.

Tools: Google co-lab, Python, Scikitlearn, Anaconda navigator

Task 13 Create a perceptron with appropriate number of inputs and outputs. Train it using fixed
incrementlearning algorithm until no change in weights is required. Output the final
weights.
Tools: Google co-lab, Python, Scikitlearn, Anaconda navigator

Part-2
Use Cases:
1. Medical Data Diagnosis
2. Biometrics pattern analysis
3. Climatic prediction
4. Credit Card Fraud Detection
5. Catching Email Spam
6. E-Commerce Product Recommendation
Total: 30 Hours
95