Course Details
Subject {L-T-P / C} : CS6218 : Machine Learning { 3-0-0 / 3}
Subject Nature : Theory
Coordinator : Prof. Anup Nandy
Syllabus
Introduction to Machine Learning: Formulation of Machine Learning Problems, Process of Classifier Design.
Mathematical Foundation: Linear Algebra, Probability Theory, Statistics, Hypothesis Testing.
Parametric Classifiers: Bayesian Decision Theory, Discriminant Functions for the Normal Density and Decision Surfaces, Errors Probability.
Parameter Estimation: Effect of Sample Size in Estimation, Maximum Likelihood and Bayesian Estimation, The Naïve Bayes Classifier.
Nonparametric Density Estimation: Parzen Density Estimate, k-Nearest Neighbor Estimation, The Nearest Neighbor Rule, Mixture Models.
Linear Classifier: Linear Discriminant Function and Decision Hyperplanes, The Perceptron Algorithm, Least Square Methods, Mean Square Estimation, Logistic Regression, Support Vector Machines, Decision Trees. K-Nearest Neighbor Rule
Non-Linear Classifier: The XOR Problem, Multilayer Neural Network, The Backpropagation Algorithm, Radial Basis Function Network, Non-Linear SVM, Probabilistic Neural Network. Hidden Markov Models.
Feature Selection and Feature Generation: Feature Selection based on Statistical Hypothesis Testing, Class Separability Measures. Dimensionality Reduction Techniques using Principal Component Analysis, Linear Discriminant Analysis, Independent Component Analysis.
Unsupervised Learning and Clustering: Unsupervised Bayesian Learning, Criterion Functions for Clustering, Sequential Clustering Algorithms, Hierarchical Clustering Algorithms, Cluster Validity.
Course Objectives
- To understand the basic building blocks and general principles that allow one to design machine learning algorithms
- To become familiar with specific, widely used machine learning algorithms
- To learn methodology and tools to apply machine learning algorithms to real data and evaluate their performance
Course Outcomes
1. Explain the principles, advantages, limitations such as overfitting and possible applications of machine learning <br />2. Identify and apply the appropriate machine learning technique to classification, pattern recognition, optimization and decision problems. <br />3. Develop an appreciation for what is involved in learning from data. <br />4. Understand how to apply a variety of learning algorithms to data. <br />5. Understand how to perform evaluation of learning algorithms and model selection.
Essential Reading
- Tom Mitchell, Machine Learning, McGraw Hill , 1997, ISBN 0-07-042807-7
- Richard O. Duda, Peter E. Hart, David G. Stork, Pattern classification, Wiley , (2nd edition). Wiley, New York, 2001
Supplementary Reading
- Christopher M. Bishop, Pattern Recognition and Machine Learning, Springer , 2011 edition
- Ian Goodfellow, Yoshua Bengio, Aaron Courville, Deep Learning, MIT Press , 2016