Parsimonious Kernel Fisher Discrimination

Abstract. By applying recent results in optimization transfer, a new algorithm for kernel Fisher Discriminant Analysis is provided that makes use of a non-smooth penalty on the coefficients to provide a parsimonious solution. The algorithm is simple, easily programmed and is shown to perform as well as or better than a number of leading machine learning algorithms on a substantial benchmark. It is then applied to a set of extreme small-sample-size problems in virtual screening where it is found to be less accurate than a currently leading approach but is still comparable in a number of cases.

• Type: Conference
• Year: 2007
• Where: IBPRIA
• Authors: Kitsuchart Pasupa, Robert F. Harrison, Peter Willett 0002
• Keywords: Algorithms, Fisher Discriminant Analysis, IBPRIA 2007, Image, Image Analysis, Non-smooth Penalty