A Comprehensive Survey on Metaheuristic Algorithm for Feature Selection Techniques

In Healthcare industries, patients’ medical records consist of a greater number of features or attributes related to certain health factors and medical terminologies. In these feature set, many attributes are either redundant one or irrelevant to the medical disease that is going to be dealt with it. These redundant and irrelevant feature sets is not going to enhance the final result at any cost and moreover it out bounds the final prediction. Feature selection is the process of selecting feature sets in a neural model that are related to a specific problem in order to aid in the inclusion of relevant features and the exclusion of irrelevant or redundant features, thereby reducing the amount of noisy data and the size of the entire dataset. In this paper, metaheuristic related techniques are taken into account and summarized about its key feature along with how it works. Also, various Feature selections related research works that have been published in last three years are highlighted with its proposed methodology, benefits and limitations. From this, it is inferred that many open challenges and enhancements are still available in the feature selection problem even though many researchers introduced their novelty in this problem.