An Evolutionary Approach of Machine Learning for Monitoring Churn Prediction of Broadband Customer

Imran Pathan et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(3), May - June 2021, 2623– 2629 2623  ABSTRACT Era of industrial Revaluation technologies and multiple end users emerging flavours of services can change the mind of end user at any time. This situation increases the demands of technological flavours. Extreme traffic of streaming, high quality of multimedia systems, flavours of services, on demand access can makes the mind diversion of humans, situation increases market competition of telecom to rendering the demanding services and that's make ruin. This situation increase such type of customer those who stop the business with entire company and deal with another company which gives its demanding services. The leading situation increases the user churn or unsatisfied users. [3, 4, 5, 6]. Presented work is an Evolutionary Approach of Machine Learning for Monitoring Churn Prediction of Broadband Customer “MCPOBBC” for telecom industries. The work ascertain the Characterization of Real world Broadband Users Data (N1=3400 and N2=6400 datasets and N3=4600 from History Log of Broadband user of Internet Service Provider: ISP) according churn prediction model to elevate the telecom industries which are providing DSL, ADSL broadband services to their customer. The whole work has been accomplished by the implementation C5-J48 decision tree algorithm on Weka 3.8 machine learning simulations and resulting predicted churn validated on CHAID, CART, and k-mean for Clustering algorithms