ABSTRACT
Feature selection using soft computing techniques is ideally used to identify and eliminate irrelevant and redundant features that do not enhance the accuracy of the prediction model. However, the process of feature selection is a daunting task, mainly due to the large search space. The domain of bionics is being explored, and metaheuristic algorithms inspired by biological processes are used to create better-designed soft computing models. This chapter compares the use of genetic algorithms combined with logistic regression and the random forest approach to optimize classification by searching for an optimal feature weightage. Selected features are used to create a neural network model and the performance metrics are assessed. It has been observed that the model constructed with the selected critical parameters exhibits a performance that is highly satisfactory and would effectively aid in the diagnosis of the disease.
