Meta-heuristic algorithms-based feature selection for breast cancer diagnosis: A systematic review

Abstract

The process of choosing a subset of significant features to be used in developing predictive models is known as feature selection. Recently, a feature selection method has been required since it is challenging to mine and convert large amounts of data into useful information. Finding victims of the feature selection approach is extremely challenging. Using the feature selection method effectively may greatly increase the classification prediction rate of various classifiers. In addition, by decreasing data size and selecting the optimal feature set in data mining and pattern recognition applications, the usage of feature selection decreases computational time and speeds up the data mining process. Therefore, feature selection algorithms have recently been used for breast cancer detection where early-stage diagnosis and treatment increase patient safety. In this paper, three meta-heuristic optimization techniques that are used in feature selection methods for breast cancer detection, namely Genetic algorithm (GA), Particle swarm optimization (PSO), and Artificial Bee Colony (ABC), will be presented. Finally, several research papers will be summarized and compared regarding the number of selected features and classification accuracy based on each algorithm.