Telecom Churn Analysis using Machine Learning in Smart Cities

Abstract

With the increase in the Telecom industry, service providers are more attentive toward the action of becoming larger or more extensive to the subscriber base. For surviving in telecom companies, the continued possession of holding customers must be a big challenge. According to consideration in the telecom environment, the market price of obtaining the new purchaser is more than holding the existing purchaser. Through collecting knowledge from the telecom industry to analyze the association of the customer whether will leave or not the company. Such types of the Decision tree and Logistic regression model have been compared on the 3334 instances of the dataset. The classification model derived from logistic regression has an accuracy of 80% and the decision tree classifier with an accuracy of 97%.