An Integrated Nature-inspired Algorithm Hybridized Adaptive Broad Learning System for Disease Classification

Classification of different deadly diseases using machine learning algorithms helps in making the health care system more robust. This, not only reduces human errors during diagnosis disease due to inexperience but also physician for taking an emergency action earlier biopsy. As genomic data undergoes through malediction excessive dimension problem, both selection remarkable genes and classification these efficiently still remain as a demanding research problem. To obtain notable features from high dimensional data, by nature-inspired algorithm, is Non-deterministic Polynomial-time (NP)-Hard Therefore, researcher can apply new algorithm solve this issue. In suggested approach, integrated natured-inspired i.e., Sine-Cosine (SC) based Monarch Butterfly Optimization (SC-MBO) merged with Adaptive Broad Learning System (ABLS) called SC-MBO-ABLS, find out most genetic classify at same time. preliminary stage, feature extraction method i.e. Kernel Fisher Score (K-FS) applied extract key gene subset. Then, extracted subset goes further execution SC-MBO-ABLS method. examine effectiveness presented method, ten datasets are considered. Here, several performance evaluators (i.e., Precision, Matthews Correlation Coefficient, Sensitivity, Kappa, Specificity, F-score) used neutral estimation approach. This model compared SC-MBO wrapped Multilayer Perceptron (SC-MBO-MLP), Extreme Machine (SC-MBO-ELM), (SC-MBO-KELM). Further, sixteen existing standard models prove supremacy Moreover, analysis Variance ANOVA test carried statistical evaluation proposed work. Eventually, according above quantitative qualitative measure, it summarized that surpasses other considering models.