CHURN FORECASTING USING DEEP LJEARNING MODEL

In the upcoming generation, people seem to be in increasing demand for the telecom sector for its products and services, and the market for the telecommunication industry is predicted to grow at a very high pace. The telecom market is modifying based on innovations and developments worldwide. For the extension and advancement of the telecommunication market, customers and their loyalty play a vital role. In this competitive world, telecommunication companies can firm their ground by building trust with customers, along with fulfilling their expectations to maintain loyalty towards them. Making the old customers stay and gaining new customers is also a crucial role for the growth of a market. But in today's time, the number of churns is high in the telecommunication industry. These churns depend on many reasons like high customer effort, ineffective marketing, or better offers from competitors. Thus, an effective way to distinguish churn from a regular customer is needed. Without any genuine analysis, the telecom industry will suffer from the loss of its customers, which the industry cannot bear. Unlike most of the research, we have used Multi-layered perceptron (MLP), for better prediction of churns by first understanding the features of the dataset and which features are more crucial in predicting the churn.