Application of nature inspired soft computing techniques for gene selection: a novel frame work for classification of cancer

A modified Artificial Bee Colony (ABC) metaheuristics optimization technique is applied for cancer classification, that reduces the classifier’s prediction errors and allows faster convergence by selecting informative genes. Cuckoo search (CS) algorithm was used in onlooker bee phase (exploitation phase) of ABC to boost performance maintaining balance between exploration exploitation ABC. Tuned using Naïve Bayes (NB) classifiers improve further accuracy model. Independent Component Analysis (ICA) dimensionality reduction. In first step, reduced dataset optimized Modified after that, second train NB classifier. Extensive experiments were performed comprehensive comparative analysis proposed with well-known metaheuristic algorithms, namely Genetic Algorithm (GA) when same framework classification six high-dimensional datasets. The comparison results showed model CS achieves highest as maximum less count selected This shows effectiveness which validated ANOVA classification.