Sparrow Search Optimization with Ensemble of Machine Learning Model for Customer Retention Prediction and Classification

Abstract

Customer Relationship Management (CRM) system is majorly used to allow organizations to obtain new customers, establishes a continuous relationship with them and rises customer retention for gaining more profits. CRM systems use machine learning (ML) algorithms for analyzing customers' behavioural and personal data to give organizations a competitive advantage by rising customer retention rates. Therefore, this study designs a Sparrow Search Optimization with Ensemble of Machine Learning Models for Customer Retention Prediction and Classification (SSOEML-CRPC) technique. The presented SSOEML-CRPC technique aims to classify the possibility of customer churn. To attain this, the SSOEML-CRPC technique follows data normalization approach to uniformly scale the data. Next, the SSO algorithm is employed for the choice of optimal features. For classifying customer churn, ensemble of ML models namely back propagation neural network (BPNN), adaptive neuro fuzzy inference system (ANFIS), and extreme gradient boosting (XGBoost) models. The experimental result analysis of the SSOEML-CRPC technique is well studied on benchmark datasets and the results can be investigated in terms of several aspects. The experimentation outcomes illustrate the better outcomes of the SSOEML-CRPC technique over other existing models.