Course Details

Subject {L-T-P / C} : EC6605 : Pattern Recognition and Applications { 3-0-0 / 3}

Subject Nature : Theory

Coordinator : Prof. Samit Ari

Syllabus

Pattern Recognition, Classification, and Description, Pattern and Feature Extraction with Examples, Training and Learning in PR Systems, Pattern Recognition Approaches The Gaussian Case and Class Dependence, Discriminant Functions, Classifier Performance, Risk, Errors, Parametric Estimation and Supervised Learning, Maximum Likelihood (ML) Estimation, Bayesian Parameter Estimation Approach, Supervised Learning Using Non-parametric Approaches, Parzen Windows, k-NN Non-parametric Estimation, Direct Classification using Training Set, Nearest Neighbor Role (NNR) Linear Discriminant Functions, Fisher’s Linear Discriminant, Discrete and Binary Classification Problems, Techniques to Directly Obtain Linear Classifiers., Formulation of Unsupervised Learning Problems, Clustering for Unsupervised Learning and Classification Syntactic Pattern Recognition Overview, Quantifying Structure in Pattern Description and Recognition, Grammar-Based Approach and Applications, Elements of Format Grammars, Recognition of Syntactic Descriptions, Parsing, The Cocke-Younger-Kasami (CYK) Parsing Algorithm, (Augmented) Transition Networks (ATNs) in Parsing, Higher Dimensional Grammars, Stochastic Grammars and Applications.

Course Objectives

Course Outcomes

A Student will learn different pattern recognition and classification techniques e.g. Parzen Windows, k-NN Non-parametric Estimation, Nearest Neighbor Role (NNR) Linear Discriminant Functions etc. and to apply these techniques for real time problems.

Essential Reading

Supplementary Reading