Course Name: CS6011 : Kernel Methods for Pattern Analysis

Description: Pattern analysis by learning from data: Risk minimization techniques for learning from data, Empirical risk minimization, Regularization, Elements of statistical learning theory, Structural risk minimization. Pattern analysis using convex optimization: Convex optimization, sequential minimal optimization, Iterative methods, Novelty detection using smallest enclosing hypersphere, Pattern classification using support vector machines, Function approximation using support vector regression . Pattern analysis using ranking, clustering and data visualization: Discovering rank relations using kernel methods, Discovering cluster structures using kernel methods, Data visualization using kernel methods. Pattern analysis using eigen-decomposition: Generalized eigen-analysis, Kernel principal component analysis, Kernel canonical correlation analysis, Kernel Fisher discriminant analysis. Construction of kernels: Properties of kernels, Basic kernels, Kernel types, Kernels for structured data, Kernels for sequential pat

Slot: A

RoomNo: CS34

Instructor: Chandra Sekhar C

Period: JAN-MAY 2013

This page was created on: Thursday 19th of September 2013 09:37:08 PM