A Study on Efficiency of Decision Tree and Multi Layer Perceptron to Predict the Customer Churn in Telecommunication using WEKA

Abstract

The Data mining is the technique to discover the knowledge which is hidden in the large data sets. It involves with different methods and algorithms to perform efficient analysis over the data sets. The classification is the technique, which is used to mine the data and helps to make the prediction about the future. Different data mining algorithms are available for classification, like C4.5, Simple Cart, NavieBayesen, Logistic Regression and Multi Layer Perceptron based on Artificial Neural Network. The main objective of this paper is to analyse the efficiency of various classification algorithms in terms of performance, accuracy and time complexity. Telecommunication churn dataset is used for the analysis. The obtained results revealed that MLP algorithm outperformed in terms of accuracy and C4.5 algorithm provides better performance in terms of time complexity.