PSO-SVM Based Algorithm for Customer Churn Prediction in the Banking Industry

Customer churn is a major concern for the banking industry and there are many methods being investigated to predict whether a customer would potentially churn. To solve this problem, one such approach is the use of machine learning techniques combined with one of the widely used models Support Vector Machine (SVM). However, SVM required the right set of hyperparameters to function optimally. This paper proposes the use of Particle Swarm Optimization (PSO) to find the optimal hyperparameters for the SVM model. The data used is obtained from Kaggle and several preprocessing techniques have been deployed to prepare the data. The results are compared with existing models that have used the same dataset. Based on the results, the proposed model obtained a significantly higher accuracy than existing SVM-based models.