Gene selection and dynamic neutrosophic cognitive map with bat algorithm (DNCM-BA) for diagnose of rheumatoid arthritis (RAs)

Abstract

Rheumatoid Arthritis (RA) is an autoimmune inflammatory rheumatic disease that has emotional impact on various body parts and tissue, principally the synovial joints. RA is a complex disease similar to many other autoimmune diseases, in which environmental aspects, genetic variants, as well as arbitrary events cooperate to activate pathological pathways. Choosing the appropriate gene for sample classification is very hard in numerous gene expression analyses in RA, in which authors attempt to find out the least probable set of genes, even now which could attain better predictive performance. On the other hand, the accuracy of classification is not up to the mark. As a result, for identifying RA disease, this research presents a gene selection as well as classification technique. Initially, with the aim of decreasing the time complexity, this disease dataset is preprocessed. Next, so as to decrease the amount of gene, the gene data is chosen from the preprocessed data by means of filter based gene selection techniques: Chi-square (CHI), Information Gain (IG), Consistency Based Subset Evaluation (CS) and Correlation Based Gene Selection (CGS). Thirdly, for the purpose of the classification of RA disease, a Dynamic Neutrosophic Cognitive Map with Bat Algorithm (DNCM-BA) is presented, that is well-suited with the medical routine and it is proposed for supporting gene expression beforehand and accurate diagnosis of RA patients. As a result, RA disease is not permitted from moving to progressive phases and the difficulty of emerging insistent as well as erosive arthritis for RA patients will get reduced. Finally, the outcome confirms that the DNCM-BA technique provides better performance while matched up with FCM–Particle Swarm Optimization (FCM-PSO), Fuzzy C Means (FCMs), Dynamic Firefly Algorithm Fuzzy C Means (DFAFCM) and Dynamic Fuzzy C Mean (DFCM) clustering algorithms in regard to precision, accurateness, recall and F-measure.