Hybrid intelligent methods for microarray data analysis

Abstract

Data produced out of microarray experiments are of great use for the physician when it is presented in a meaningful manner. This paper proposes hybrid intelligent methods for addressing the challenges in analyzing the microarray data. The concept of fuzzy and rough set is hybridized with FInformation (FRFI) for gene selection. An optimal fuzzy logic based classifier (FLC) is developed for sample classification using a hybrid Genetic Swarm Algorithm (GSA). Detailed experiments are conducted using microarray data related to Cancer and Rheumatoid Arthritis. From the simulation study, it is found that the proposed FRFI-FLC-GSA produces compact classification system with reasonably good informative genes that can be used for disease diagnosis.