Optimal Feature Selection using Enhanced Bacterial Foraging Optimization for Healthcare Application Based Big Data

An important and complex set of data is referred as big data, which is contains big amount of data. The big data is used to collect, storage, and analysis. Numerous large datasets feature deceptive and unnecessary attributes, leading to confusion in the classification process. Additionally, the exhaustive search within the solution space results in prolonged processing times for datasets. The disadvantages associated with high-dimensional datasets encompass prolonged model building times, the presence of redundant information, and a diminished quality that significantly hampers the process of data analysis. Addressing this issue involves employing feature selection as a primary pre-processing step. This necessitates selecting a subset of features from the large dataset to enhance better classification. So, in this work, an enhanced Bacterial Foraging Optimization (EBFO) method is used for optimal feature selection. Besides, the search ability of BFO is enhanced by adapting opposition-based learning (OBL) strategy. Subsequently, the chosen features are input into the radial basis function neural network (RBFNN). Then, different datasets were employed for the model's performance assessment. The effectiveness of the proposed model is analyzed in terms of model accuracy.