KAKATIYA UNIVERSITY WARANGAL

Under Graduate Courses (Under CBCS with effect from Academic Year 2022-2023 onwards)
B.Sc. DATA SCIENCE
II Year: Semester-IV

Paper – IV: Machine Learning

Objectives: The main objective of this course is to teach the principles and foundations of
machine learning algorithms

Outcomes:
At the end of the course the student will be able to understand
 Basics of Machine Learning and its limitations
 Machine Learning Algorithms: supervised, unsupervised, bio-inspired
 Probabilistic Modeling and Association Rule Mining

Unit-I

Introduction: What does it mean to learn, Some canonical Learning Problems, The Decision
Tree Model of Learning, Formalizing the Learning Problem ID3 Algorithm [Reference1, 2]
Limits of Learning: Data Generating Distributions, Inductive Bias, Not Everything
islearnable, Under fitting and Overfitting, Separation of training and test Data, Models,
parameters and Hyperparameters, Real World Applications of Machine Learning Geometry
and Nearest Neighbours: From Data to Feature Vectors, k-Nearest Neighbours, Decision
Boundaries, k-means Clustering, High Dimensions [Reference 1]

Unit-II
The Perceptron: Bio-inspired Learning, The Perceptron Algorithm, Geometric
Interpretation, Interpreting Perceptron Weights, Perceptron Convergence and Linear
Separability, Improved Generalization, Limitations of the Perceptron
Practical Issues: Importance of Good Features, Irrelevant and Redundant Features, Feature
Pruning and Normalization, Combinatorial Feature Explosion, Evaluating Model
Performance, Cross Validation, Hypothesis Testing and Statistical Significance, Debugging
Learning Algorithms, Bias Variance tradeoff
Linear Models: The Optimization Framework for Linear Models, Convex Surrogate Loss
Functions, Weight Regularization, Optimization and Gradient Descent, Support Vector
Machines [Reference 1]

Unit-III
Probabilistic Modelling: Classification by Density Estimation, Statistical Estimation, Naïve
Bayes Models, Prediction [Reference 1]
Neural Networks: Bio-inspired Multi-Layer Networks, The Back-propagation Algorithm,
Initialization and Convergence of Neural Networks, Beyond two layers, Breadth vs Depth,
Basis Functions [Reference 1]
Unit IV
Unsupervised Learning: Clustering Introduction, Similarity and Distance
Measures,Agglomerative Algorithms, Divisive Clustering, Minimum Spanning Tree
[Reference 2]
Association Rules: Introduction, large Itemsets, Apriori Algorithm [Reference 2]
References:
1. A Course in Machine Learning (CIML). Hal Daume III, 2017 (freely available online)
http://ciml.info/
2. Data Mining: Introductory and Advanced Topics. Margaret H Dunham, Pearson
Education, 2003
Suggested Reading:
3. Hands on Machine Learning with SciKit-Learn, Keras and Tensor Flow. AurélienGéron.
O’Reily, 2019
4. Machine Learning with Python Cookbook. Chris Albo, O’Reily, 2018
5. Introduction to Machine Learning with Python: A guide. Andreas C Miller, Sarah Guido.
O’Reily, 2017
 KAKATIYA UNIVERSITY WARANGAL
Under Graduate Courses (Under CBCS AY: 2021-2022 on words)
B.Sc. DATA SCIENCE
II Year: Semester-IV

Practical- 4: Machine Learning (Lab)

Objective:
The main objective of this laboratory is to put into practice the various machine learning
algorithms for data analysis using Python and Weka.

ML Toolkits
Students are expected to learn
1. Scikit-learn(https://scikit-learn.org/) an open source machine learning Python library that
supports supervised and unsupervised learning. It also provides various tools for model
fitting, data preprocessing, model selection and evaluation, and many other utilities.
2. Weka (http://www.cs.waikato.ac.nz/ml/weka/)is another widely used ML toolkit.

Datasets
1. The sklearn datasets package embeds small toy datasets. It includes utilities to load
these datasets. It also includes methods to load and fetch popular reference datasets
and features some artificial data generators. Students are expected to study and make
use of these datasets
2. Weka also has provides various data sets.

References:
1. Scikit-learn user guide.https://scikit- learn.org/stable//_downloads/scikit- learn-docs.pdf
2. Ian Witten, Eibe Frank, and Mark Hall, Chris Pal. DATA MINING: Practical Machine
Learning Tools and Techniques, 4th Edition. Morgan Kaufmann.
Exercises
1. Write a Python program using Scikit-learn to split the iris dataset into 70% train data and
30% test data. Out of total 150 records, the training set will contain 120 records and the
test set contains 30 of those records. Print both datasets
2. Write Python program to use sklearn’s Decision Tree Classifier to build a decision tree
for the sklearn’s datasets. Implement functions to find the importance of a split (entropy,
information gain, gini measure)
3. Write a Python program to implement your own version of the K-means algorithm. Then
apply it to different datasets and evaluate the performance.
4. Design a perceptron classifier to classify handwritten numerical digits (0-9). Implement
using scikit or Weka.
5. Write a Python program to classify text as spam or not spam using the Naïve Bayes
Classifier
6. Use WEKA and experiment with the following classifiers: Association Rule Mining
(Apriori), Agglomerative and Divisive Clustering.