Boosting Ant Colony Optimization with Reptile Search Algorithm for Churn Prediction

The telecommunications industry is greatly concerned about customer churn due to dissatisfaction with service. This has started investing in the development of machine learning (ML) models for prediction extract, examine and visualize their customers’ historical information from a vast amount big data which will assist further understand needs take appropriate actions control churn. However, high-dimensionality large influence on performance ML model, so feature selection (FS) been applied since it primary preprocessing step. It improves model’s by selecting salient features while reducing computational time, can this sector building effective models. paper proposes new FS approach ACO-RSA, that combines two metaheuristic algorithms (MAs), namely, ant colony optimization (ACO) reptile search algorithm (RSA). In developed ACO-RSA approach, an ACO RSA are integrated choose important subset prediction. evaluated seven open-source datasets, ten CEC 2019 test functions, its compared particle swarm (PSO), multi verse optimizer (MVO) grey wolf (GWO), standard RSA. According results along statistical analysis, superior other competitor most datasets.