Research on mixed-strategy based grey wolf optimization for gene expression classification

With the advantages of easy implement and few adjustment parameters, grey wolf optimization (GWO) performs well in solving global problems. However, original GWO can suffer from slow convergence speed plunge into local optimums dealing with complex To address these issues, improvement mixed-strategy based (MGWO) is proposed this paper. Firstly, we present chaos strategy combination quasi-opposition learning for generating more high-quality population to improve algorithm. Then, embed sine cosine mechanism jump out optimums. We utilize interaction better balance exploration exploitation capabilities Finally, use benchmark functions testing performance Experimental results demonstrate that algorithm search accuracy compared other state-of-the-art algorithms. Simultaneously, verify further algorithm, classification problems on gene expression datasets are also evaluated. Both continuous binary versions MGWO respectively used completing parameters feature selection. show has than others terms Sensitivity, Specificity, G-mean, Accuracy F-measure.