A New Dimensional Reduction Based on Cuttlefish Algorithm for Human Cancer Gene Expression

Abstract

Currently, the main problem in DNA Microarray is classification due to the thousands of numbers of genes together, and this huge number of genes can make the classification task very difficult. Therefore, feature selection is a very important task for gene classification. This paper presents a new model which uses a Cuttlefish Algorithm (CFA) to select the most informative features, while K-Nearest Neighbor (KNN) is used to measure the quality of the selected features that are produced by the CFA. Eight datasets are used to evaluate the performance of the proposed model and compared with the performance of four well-known existing classification techniques such as KNN, DT, Hidden Markov models (HMM), and SVM. The obtained results show that the proposed technique outperforms these existing techniques in five datasets among eight datasets.