SVM BASED CHURN ANALYSIS FOR TELECOMMUNICATION

Data mining is a technical analysis for the massive database with meaning full output. Churn analysis is best applications for the data mining able to determine the behavior of the customer in changing different services. The available marketing tool has the limitation to predict the changing behavior for the customer. This paper discuss about the churn analysis in telecommunication sector for the 2019 year Q2 period. Machine Learning is an advanced development in data mining to extract the features from large quantities of data. The paper discuss about supervised machine learning model. The supervised model designed by support vector machine (SVM) classification steps for two group separation of churn customer and non churn customer. The model aim to analysis total number of subscriber in voice call routing along the period. The churn analysis can predict the churn rate and the probability for month and the year. The proposed method, classify the customer from churn and nonchurn by SVM increase in accuracy for the existing system. The proposed work implemented using MALAB R 2014b simulation software and the results were discussed.