PSTAT262FE

Provides an introduction to core linear and non-linear feature extraction methods that form the basis of the mathematical treatment of modern multivariate analysis in machine learning practice and kernel methods. The topics covered will include aspects of kernel theory, efficient and robust computation with large Gram matrices, kernel construction, multivariate methods such as Principle Component Analysis (PCA), kernel PCA, Independent Component Analysis (ICA), Partial Least Squares (PLS) and kernel PLS, Canonical Correlation Analysis (CCA) and Kernel CCA, Fisher Discriminant Analysis and kernel FDA. Regularization methods in feature extraction, compressive sampling and random projections.

May be repeated for credit.