A comprehensive survey of feature selection techniques based on whale optimization algorithm

Machine learning and data mining rely on feature selection to reduce the dimension of data and increase the performance of algorithms. As a result of such a large search space, feature selection is a challenging task. Recently, evolutionary techniques have been gaining a lot of attention and showing some promise for solving feature selection problems. Recent studies have shown that Whale Optimization Algorithm (WOA) is widely used in various fields (e.g., data mining, machine learning, and cloud computing). Motivated by the extensive research efforts in the feature selection and WOA, we present a review of high-quality articles related to WOA-based feature selection algorithms published between 2017 and 2023. This paper discusses and compares WOA-based feature selection schemes based on merits and demerits, evaluation techniques, simulation environments, and important parameters. We begin by introducing feature selection process, and concepts of metaheuristic followed by their surveys. This study summarizes several domains where WOA is used and explains different types of features. Moreover, it categorizes the variants of WOA based on their learning process, parameter tuning, binary/discrete, and hybridization. According to the investigation results, few variations of WOA add new parameters or operators to the original. In addition, 60% of feature selection algorithms based on WOA focus on improving learning process. Finally, current issues and challenges are also discussed to identify future research areas.