Welcome to the InfoSci Platform
Reference Hub3
A Hybrid Kernel Extreme Learning Machine and Improved Cat Swarm Optimization for Microarray Medical Data Classification

A Hybrid Kernel Extreme Learning Machine and Improved Cat Swarm Optimization for Microarray Medical Data Classification

S. Chakravarty, R. Bisoi, P. K. Dash
Copyright: © 2016 |Volume: 7 |Issue: 3 |Pages: 30
ISSN: 1942-3594|EISSN: 1942-3608|EISBN13: 9781466690820|DOI: 10.4018/IJAEC.2016070104
Cite Article Cite Article

MLA

Chakravarty, S., et al. "A Hybrid Kernel Extreme Learning Machine and Improved Cat Swarm Optimization for Microarray Medical Data Classification." IJAEC vol.7, no.3 2016: pp.71-100. http://doi.org/10.4018/IJAEC.2016070104

APA

Chakravarty, S., Bisoi, R., & Dash, P. K. (2016). A Hybrid Kernel Extreme Learning Machine and Improved Cat Swarm Optimization for Microarray Medical Data Classification. International Journal of Applied Evolutionary Computation (IJAEC), 7(3), 71-100. http://doi.org/10.4018/IJAEC.2016070104

Chicago

Chakravarty, S., R. Bisoi, and P. K. Dash. "A Hybrid Kernel Extreme Learning Machine and Improved Cat Swarm Optimization for Microarray Medical Data Classification," International Journal of Applied Evolutionary Computation (IJAEC) 7, no.3: 71-100. http://doi.org/10.4018/IJAEC.2016070104

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

This paper presents the pattern classification of the binary microarray gene expression based medical data using extreme learning machine (ELM) and its variants like on-line sequential ELM (OSELM) and kernel based extreme learning machine (KELM). In the KELM category two variants namely the wavelet based kernel (WKELM) extreme learning machine and radial basis kernel extreme learning machine (RKELM) along with support vector machine (SVMRBF) and support vector machine polynomial (SVMPoly) are used to classify microarray medical datasets. Further to reduce the high dimensionality of Microarray medical datasets giving rise to high number of gene expression and small sample sizes, a modified evolutionary cat swarm optimization (MCSO) technique is adopted. The efficiency of the proposed algorithm is verified using a set of performance metrics for four binary medical datasets belonging to breast cancer, prostate cancer, colon tumor, and leukemia, respectively.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.