Pattern Recognition

What you’ll study

Introduction to pattern recognition — statistical, structural and hybrid classification methods, and applications to character recognition and medical imaging; feature detection and classification; a review of probability and linear algebra. Bayesian decision making — linear discriminants, separability, multi-class discrimination, quadratic classifiers, the Fisher discriminant and sufficient statistics; coping with missing or noisy features. Parametric and non-parametric estimation — maximum-likelihood and Bayesian parameter estimation, MAP, density estimation and Parzen estimation; non-parametric classification and k-nearest-neighbour classification. Template-based recognition, eigenvector analysis and feature extraction. Clustering and unsupervised learning — vector quantization, K-means and expectation-maximization, and neural nets. Sequence analysis — hidden Markov models, the Viterbi and Baum-Welch algorithms, linear dynamical systems, Kalman filtering, Bayesian networks, decision trees and multi-layer perceptrons.