Ensemble classifier based big data classification with hybrid optimal feature selection

Abstract

Big data is turning out to be well-liked and enviable amongst numerous users for storing, analyzing, and handling larger quantities of data. Nevertheless, clustering these larger data has turned into more multifaceted owing to its data size. A number of machine learning (ML) approaches have been developed recently to extract information from Big Data. These existing techniques, on the other hand, do not meet the accuracy criterion. Proposed LDA, PCA, and LSR-based features are first calculated. The optimal features are then chosen using a new SSI-CSA model. These optimal features are then classified via ensemble classifier (EC) that includes SVM, RF, DT and NN and the precise classified outcomes are obtained. This work employs the Shark Smell Integrated Cat Swarm Algorithm (SSI-CSA) model for precise feature selection. In the end, the improvement of deployed scheme is confirmed regarding diverse metrics like FNR, MCC, and accuracy and so on.