Customer Churn Prediction in Telecom Industry Using Deep Learning Techniques

Abstract

In recent times, the telecommunications request has beenveritably competitive. The cost of retaining telecom guests is lower thanattracting new guests. A telecom company must understand client churn through client relationship management. Therefore, CRM analyzersare demanded to prognosticate which guests will change. In this design, aclient abandonment vaticination model was proposed that uses the Decision Treealgorithm, Random timber algorithm, and Deep literacy algorithm to identify thechurn guests. In Deep Literacy, a Multilayer Perceptron Neural Network has beenused to produce the vaticination model. The performance of all the algorithmswill be compared and estimated in terms of delicacy. The advanced performancevaticination model can be used in the telecom sphere for prognosticatingwhether the guests will churn or not. By knowing the significant churn factorsfrom client data, CRM can ameliorate productivity, recommend applicableelevations to the group of likely churn guests grounded on analogous getspatterns, and exorbitantly ameliorate marketing juggernauts of the company.