Dynamic Population-Based Jaya Algorithm for Genetic Biomarker Classification

Cancer is among the leading causes of death worldwide. If we can predict an individual’s risk of developing cancer, then steps can be taken to minimize the risk of developing a malignancy. One of the useful techniques is to identify the genetic biomarkers that are responsible for causing cancer. Detection and classification of these biomarkers require either biopsy computer-aided imaging or microarray data analysis. This paper proposes a new technique for the detection and classification of biomarkers using microarray data analysis. The work is based on a hybrid method where gene selection is carried out using minimum redundancy maximum relevance (mRMR) followed by a dynamic population-based Jaya algorithm. Here logistic map is employed for faster convergence. The majority of the existing evolutionary algorithms (EAs) are based on fixed population size which increases computation time. Therefore, a dynamic population-based evolutionary algorithm is employed here. Specifically, the goal is to investigate if there is a significant advantage of reducing the population size with the increase in iterations. To validate the performance of the proposed method, three state-of-the-art algorithms are selected for comparison over three benchmark datasets. The results of our experiments prove that our proposed method performs better than fixed population-based methods.