MIT Department of Media Arts and Sciences and MIT Department of Civil and Environmental Engineering
MAS.622J/1.126J: Pattern Recognition and Analysis (Autumn 2004 Teaching Assistant)
Instructor: Rosalind Picard
Topics include: introduction to pattern recognition, feature detection, classification;
review of probability theory, conditional probability and Bayes rule;
random vectors, expectation, correlation, and covariance;
review of linear algebra, and linear transformations;
decision theory, ROC curves, and likelihood ratio test;
linear, quadratic, and Fisher discriminants;
sufficient statistics and coping with missing or noisy features;
template-based recognition, eigenvector analysis, and feature extraction;
training methods, maximum likelihood, and Bayesian parameter estimation;
linear discriminant/perceptron learning, optimization by gradient descent, and SVM;
k-nearest-neighbor classification;
non-parametric classification, density estimation, and Parzen estimation;
unsupervised learning, clustering, vector quantization, and k-means;
mixture modeling and optimization by Expectation-Maximization;
hidden Markov models, Viterbi algorithm, and Baum-Welch algorithm;
linear dynamical systems and Kalman filtering and smoothing;
Bayesian networks and independence diagrams; decision trees and multi-layer perceptrons;
and combination of multiple classifiers as "committee machines".
[official course page]
|
|
|