Pattern Recognition

Pre-Requisite : EC301
Contact Hours and Credits : ( 3 -0- 0 ) 3

Objective : 

To develop the mathematical tools required for the pattern recognition.

Topics Covered :

Fundamental concepts and blocks of a typical pattern recognition system. Decision functions- role and types, pattern and weight space, properties and implementation of decision functions.

Feature identification, selection and extraction. Distance measures, clustering transformation and feature ordering, clustering in feature selection, feature selection through maximization and approximations.

Pattern classification by distance functions. Clusters and cluster seeking algorithms. Pattern classification by likelihood functions. Baye’s classifier and performance measures.

Artificial neural network model, Neural network-based pattern associators, Feed forward networks and training by back-propagation- ART networks.

Applications of statistical and neural network – based pattern classifiers in speech recognition, image recognition and target recognition.

Course Outcomes :

On the successful completion of this course Student are able 

  • CO1: Summarize the various techniques involved in pattern recognition
  • CO2: Categorize the various pattern recognition techniques into supervised and unsupervised.
  • CO3: Illustrate the artificial neural network based pattern recognition
  • CO4: Discuss the applications of pattern recognition in various applications

Text Books:

  • J.I. Tou & R.C. Gonzalez, Pattern Recognition Priciples, Addition-Wesley.
  • R. Schalkoff, Pattern Recognition - Statistiucal, Structural and Neural Approaches, John Wiley, 1992.

Reference Books:

  • P.A. Devijer & J. Kittler, Pattern Recognition - A Statistical Approach, Prentice-Hall.
  • Christopher. M. Bishop, 'Pattern recognition and machine learning, Springer, 2006.