[Compstats] Compstats seminar Wed @ 2:30pm: Hadi Zarkoob

Pascal Poupart ppoupart at cs.uwaterloo.ca
Tue Apr 20 11:45:38 EDT 2010

Title: Fast Feature Selection for Gene Expression Data via Hilbert-Schmidt
Independence Criterion.

Speaker: Hadi Zarkoob
Date: April 21
Time: 2:30pm
Location: AI seminar room (DC2306C)

Abstract: A novel feature selection technique for microarray gene
expression data is proposed. It is based on the Hilbert-Schmidt
independence criterion, and partly motivated by Rank-One Downdate
(R1D) and the Singular Value
Decomposition (SVD). The algorithm selects a small set of genes such that
the response variable depends mainly on this subset, at the exclusion of the
rest of the genes. The algorithm is computationally very fast and scalable
to large data sets, and it does not require the number of important genes as
an explicit input. Experimental results of the proposed technique are
presented on some synthetic and well-known microarray data sets.

Pascal Poupart
Associate Professor
David R. Cheriton School of Computer Science
University of Waterloo
200 University Avenue West
Waterloo, Ontario
Canada N2L 3G1
Web: http://www.cs.uwaterloo.ca/~ppoupart
Email: ppoupart at cs.uwaterloo.ca 
Telephone: 1-519-888-4567x36239 
Fax: 1-519-885-1208

More information about the Compstats mailing list