Accuracy Measure of Customer Churn Prediction in Telecom Industry using Adaboost over Random Forest Algorithm
DOI:
https://doi.org/10.47750/pnr.2022.13.S04.178Keywords:
Customer churn, Novel Adaboost Algorithm, Random Forest algorithm, Machine Learning, Telecom Industry, Data AnalyticsAbstract
Aim: To increase the customer churn prediction model accuracy in the telecom industry using Adaboost over Random Forest
Algorithm.
Materials and methods: Adaboost algorithm and Random Forest algorithm with sample size (N=10) is executed with varying training and testing splits for predicting the accuracy for customer churn prediction and achieved the G power of 75% and threshold 0.000 and confidence interval 95%. The performance of the model is calculated based on their accuracy rate using the customer churn dataset. Results and Discussion: The customer churn prediction model has attained an accuracy of 90% using Novel Adaboost algorithm and 81% using Random Forest algorithm. There exists a statistical difference between Novel Adaboost and Random Forest (p=0.023) where p <0.005 .
Conclusion: Prediction of customer churn using the Novel Adaboost algorithm results significantly greater than the Random Forest algorithm with improved accuracy.
