CSDL Home IEEE/ACM Transactions on Computational Biology and Bioinformatics 2011 vol.8 Issue No.04 - July/August
Issue No.04 - July/August (2011 vol.8)
David W. Bulger , Macquarie University, Sydney
Tim Peters , Macquarie University, Sydney
Jean Yee Hwa Yang , University of Sydney, Sydney
David Ma , St. Vincent's Hospital and St. Vincent's Centre for Applied Medical Research, Sydney
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TCBB.2011.30
Current feature selection methods for supervised classification of tissue samples from microarray data generally fail to exploit complementary discriminatory power that can be found in sets of features [CHECK END OF SENTENCE]. Using a feature selection method with the computational architecture of the cross-entropy method [CHECK END OF SENTENCE], including an additional preliminary step ensuring a lower bound on the number of times any feature is considered, we show when testing on a human lymph node data set that there are a significant number of genes that perform well when their complementary power is assessed, but "pass under the radar” of popular feature selection methods that only assess genes individually on a given classification tool. We also show that this phenomenon becomes more apparent as diagnostic specificity of the tissue samples analysed increases.
Feature selection, microarray, data mining, genetic interdependence, lymphoma.
David W. Bulger, Tim Peters, Jean Yee Hwa Yang, David Ma, "Two-Step Cross-Entropy Feature Selection for Microarrays—Power Through Complementarity", IEEE/ACM Transactions on Computational Biology and Bioinformatics, vol.8, no. 4, pp. 1148-1151, July/August 2011, doi:10.1109/TCBB.2011.30