Bio-Inspired Approach to Extend Customer Churn Prediction for the Telecom Industry in Efficient Way

Abstract Churn prevention has always been a top priority in business retention. The significant problem of customer churn was confronted by the telecommunications industry due to saturated markets, harsh competition, dynamic criteria, as well as the launch of new tempting offers. By formalizing the telecom industry's problem of churn prediction as a classification task, this work makes a contribution to the field. To effectively track customer churn, a churn prediction (CP) model is needed. Therefore, using the deep learning model known as the reformatted recurrent neural network in conjunction with the Elephant herding optimization (EHO) method, this work provides a novel framework to forecast customer turnover (R-RNN). EHO is a meta-heuristic optimization algorithm that draws inspiration from nature and is based on the herding behaviour of elephants. The distance between the elephants in each clan in relation to the location of a matriarch elephant is updated by EHO using a clan operator. For a wide range of benchmark issues and application domains, the EHO approach has been shown to be superior to several cutting-edge meta-heuristic methods. In order to classify the Churn Customer (CC) and a regular customer, RRNN is modified. This improved EHO effectively optimises the specific RNN parameters. If a client churns as a result, network usage is examined as a retention strategy. However, this paradigm does not take into account the number of consumers who leave based on how often they use their local networks. The results of the simulation and performance metrics-based comparison are assessed to show that the newly proposed technique can identify churn more successfully than pertinent techniques.