Influential Gene Selection From High-Dimensional Genomic Data Using a Bio-Inspired Algorithm Wrapped Broad Learning System

The classification of high dimensional gene expression/ microarray data always plays an important role in various disease diagnoses and drug discovery. To avoid the curse dimensionality, selection most influential genes remains a challenging task for researchers machine learning field. As extraction features by bio-inspired algorithm is non-deterministic polynomial-time (NP-Hard) task, possibility applying new there. In this suggested work, recently developed algorithm, Monarch Butterfly Optimization (MBO), wrapped with Broad Learning System (BLS), called MBO-BLS, to choose classify simultaneously. first stage, pre-selection method (Relief) used select feature subset. Then, selected subset undergoes further execution MBO-BLS model. estimate efficacy presented model, six cancerous datasets are taken. Here, sensitivity, specificity, precision, F-score, Kappa, MCC measures impartial comparison. Further, prove supremacy method, basic BLS, Genetic Algorithm BLS (GA-BLS), Particle Swarm (PSO-BLS), existing ten models taken Moreover, examine designed model statistically, Analysis variance (ANOVA) test also performed here. From above qualitative quantitative analysis, it concluded that proposed outclasses other considering models.