Analysis of Customer Churn Prediction
in Telecom Sector Using CART
Algorithm
Sandeep Rai, Nikita Khandelwal and Rajesh Boghey
Abstract Predicting client churn in telecommunication industries becomes the most
significant topic for analysis in recent years. Because its helps in detecting which
customer are likely to change or cancel their subscription to a service. Analysis of
information that is extracted from telecommunication companies will help to seek out
the explanations of client churn and also uses the knowledge to retain the purchasers.
Thus, predicting churn is extremely necessary for telecommunication firms to retain
their customers. During this paper, we have designed the classification model using
call tree, evaluated the performance measures, and compared its performance with
logistic regression model.
Keywords Classification · Churn prediction · Telecom data · Logistic regression
model · Customer retention · CART algorithm
1 Introduction
Data mining strategies lie at the intersection of computing, statistics, and machine
learning info systems. Data processing techniques help in building the prediction
models to get future developments and actions permitting the organizations to require
good selections derived from the data from knowledge [1].
Churn prediction is associate application of client performance in data processing.
Churn [2] could be a key issue sweet-faced through associate enterprise associated
denoted the value of extending a replacement client is almost five times more than the
value of maintaining an recent client. As a result of the fight of the enterprise market
S. Rai (B ) · N. Khandelwal · R. Boghey
Department of Computer Science and Engineering, Technocrats Institute of Technology
(Excellence), Bhopal, India
e-mail: sandtec@gmail.com
N. Khandelwal
e-mail: nikitakhandelwal0000@gmail.com
R. Boghey
e-mail: rajeshboghey@gmail.com
© Springer Nature Singapore Pte Ltd. 2020
A. K. Luhach et al. (eds.), First International Conference on Sustainable Technologies
for Computational Intelligence, Advances in Intelligent Systems and Computing 1045,
https://doi.org/10.1007/978-981-15-0029-9_36
457