Abstract—This work presents the method to classify the gene expression cancer data –Microarray data. The proposed method combines two techniques: classification and feature selection. The classification technique used in this work is Genetic Algorithm (GA) and the feature selection technique is Signal-to-Noise Ratio (SNR). Lymphoma and Leukemia datasets are used to test the performance of the proposed method and 10-Folds cross validation technique is applied to report the experimental results in term of classification accuracy. The results show that the proposed method yields the best result comparing with the simple GA-based classifier in both classification accuracy and the number of generations to found the solutions. Additionally, the results are compared to the other classification and feature selection techniques reported in the literature and it is found that the proposed method achieves a good result, especially, in the Lymphoma dataset the proposed method is the best. 

Index Terms—Microarray Data; Genetic Algorithm; Signal-to-Noise Ratio; Classification; Feature Selection

Cite: Supoj Hengpraprohm, "GA-Based Classifier with SNR Weighted Features for Cancer Microarray Data Classification," International Journal of Signal Processing Systems, Vol. 1, No. 1, pp. 29-33,  June 2013.doi: 10.12720/ijsps.1.1.29-33