Submit Manuscript  

Article Details


Hybrid Feature Selection Algorithm mRMR-ICA for Cancer Classification from Microarray Gene Expression Data

Author(s):

Shuaiqun Wang*, Wei Kong, Aorigele, Jin Deng, Shangce Gao and Weiming Zeng  

Abstract:


Microarray gene expression data is crucial to the detection of cancer and has gained more attention. Because high-dimension gene expression datasets contain a lot of redundant information which makes it difficult for cancer classification. It is very important for researchers to find appropriate ways to select informative genes for better identification of cancer. Imperialist competition algorithm (ICA) is a relatively new meta-heuristic method and has great advantages in solving many combinatorial optimization problems. For high-dimension microarray expression datasets with a large number of genes, single evolutionary algorithm is inefficient for feature selection. Furthermore, excessive genes lead to poor classification accuracy for classifier. Therefore, we present a hybrid feature selection method mRMR-ICA which combines minimum redundancy maximum relevance (mRMR) with ICA for cancer classification in this paper. The presented algorithm mRMR-ICA utilizes mRMR to delete redundant genes and provide the small datasets for ICA. It will use support vector machine (SVM) to evaluate the classification accuracy for feature genes. Ten benchmark microarray gene expression datasets are used to test the performance of mRMR-ICA. Experimental results including the accuracy of cancer classification and the number of informative genes are improved for mRMR-ICA compared with the original ICA and other evolutionary algorithms.

Keywords:

Imperialist competition algorithm, Minimum redundancy maximum relevance, Support vector machine, Microarray gene expression data, Feature selection, Cancer classification.

Affiliation:

College of Information Engineering, Shang Maritime University, Shanghai, College of Information Engineering, Shang Maritime University, Shanghai, Faculty of Engineering, University of Toyama, Toyama, College of Information Engineering, Shang Maritime University, Shanghai, Faculty of Engineering, University of Toyama, Toyama, College of Information Engineering, Shang Maritime University, Shanghai



Full Text Inquiry