Parsimonious Kernel Fisher Discrimination


Pasupa, Kitsuchart, Harrison, Robert F. and Willett, Peter (2007) Parsimonious Kernel Fisher Discrimination. At Proceeding of 3rd Iberian Conference on Pattern Recognition and Image Analysis (IbPria'2007), Girona, Spain, 06 - 08 Jun 2007. Springer Berlin / Heidelberg, 531-538.

Download

[img] PDF - Accepted Version
Download (176Kb)

Description/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.

Item Type: Conference or Workshop Item (Speech)
Additional Information: Event Dates: 6-8 June 2007
ISBNs: 9783540728467
ISSNs: 0302-9743
Related URLs:
Divisions: Faculty of Physical and Applied Science > Electronics and Computer Science
Item ID: 266582
Date Deposited: 19 Aug 2008 22:27
Last Modified: 02 Mar 2012 11:40
Contributors: Pasupa, Kitsuchart (Author)
Harrison, Robert F. (Author)
Willett, Peter (Author)
Date: 6 June 2007
Additional Information: Event Dates: 6-8 June 2007
Status: Published
Publisher: Springer Berlin / Heidelberg
Further Information:Google Scholar
ISI Citation Count:1
URI: http://eprints.soton.ac.uk/id/eprint/266582

Actions (login required)

View Item View Item