Customer Churn Analysis in Telecom Using Machine Learning Techniques

Mittal, Manish Kumar (2022) Customer Churn Analysis in Telecom Using Machine Learning Techniques. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (2MB) | Preview

Abstract

Telecommunication industry is among few industries with high technological dependency. Companies are struggling to retain their performance. For this, customer churn prediction becomes crucial way to predict customer’s information for decisions. Therefore, the study has focused on customer churn prediction the efficient role of machine learning and hybrid modelling techniques. Gradient boosting, random forest, decision tree and logistic regression has been used as machine learning techniques along with hybrid modelling. RandomizedsearchCV was used to improve gradient boost performance. Synthetic Minority Oversampling Technique was used to improve Knowledge Discovery in Databases for better results. The study results are evaluated based on confusion matrix and compared based on accuracy, precision, recall and f1 score. Gradient boosting outperformed all other models by achieving 96.81% of accuracy.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Customer Churn; Classification; Decision Tree; Random Forest; Logistic Regression; Gradient Boosting
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HF Commerce > Marketing > Consumer Behaviour Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning T Technology > TK Electrical engineering. Electronics. Nuclear engineering > Telecommunications
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	23 Feb 2023 13:09
Last Modified:	02 Mar 2023 08:53
URI:	https://norma.ncirl.ie/id/eprint/6230

Actions (login required)

View Item