Gene Selection and Classification of Microarray Data Using Convolutional Neural Network

Gene expression profiles could be generated in large quantities by utilizing microarray techniques. Currently, the task of diagnosing diseases relies on gene expression data. One of the techniques which helps in this task is by utilizing deep learning algorithms. Such algorithms are effective in the identification and classification of informative genes. These genes may subsequently be used in predicting testing samples' classes. In cancer identification, the microarray data typically possesses minimal samples number with a huge feature collection size which are hailing from gene expression data. Lately, applications of deep learning algorithms are gaining much attention to solve various challenges in artificial intelligence field. In the present study, we investigated a deep learning algorithm based on the convolutional neural network (CNN), for classification of microarray data. In comparison to similar techniques such as Vector Machine Recursive Feature Elimination and improved Random Forest (mSVM-RFE-iRF and varSeIRF), CNN showed that not all the data have superior performance. Most of experimental results on cancer datasets indicated that CNN is superior in terms of accuracy and minimizing gene in classifying cancer comparing with hybrid mSVM-RFE-iRF.