loading page

Performance analysis of Customer Attrition Prediction using Logistic Regression and K-Means Clustering.
  • +1
  • Akshara sri L,
  • Aameer Khan S,
  • Nithisshkrishna KS,
  • Anitha R
Akshara sri L
Rajalakshmi Engineering College

Corresponding Author:[email protected]

Author Profile
Aameer Khan S
Rajalakshmi Engineering College
Author Profile
Nithisshkrishna KS
Rajalakshmi Engineering College
Author Profile
Anitha R
Rajalakshmi Engineering College
Author Profile

Abstract

The quick development of technological infrastructure has significantly altered how organizations carry on with work. Subscription-based services are among the results of continuous digitization and customer attrition has become a major problem and a threat to all firms. Customer attrition, alternatively referred to as customer turnover, refers to the departure of customers over time, which is facing challenges in various business industries. To increase the customer retention percentage and for the overall profitability of the industry, customer churn must also be reduced. When organizations recognize client attrition, they can take proactive measures to keep customers. Customer attrition is a terminology adopted by different organizations to encapsulate the defection of clients or subscribers to any phenomenon. With the use of big data architecture, notably Spark, this study presents a web application for extracting telecom data. It uses machine learning algorithms like Logistic regression and K-means clustering, evaluates the performance of the models, and combines hard and soft data in order to predict customer churn more accurately. In addition, label selection will be carried out by assessing each feature’s impurity score independently, and cluster classification will be carried out to select the best cluster based on its metrics. The study concentrates on the crucial machine learning methods for calculating client churn. This can include improving customer service, offering loyalty programs, or adjusting pricing strategies.