Intellectual Property India Publishes Patent Application for 'Telecom Customer Churn Prediction And Analysis' Filed by Tatikonda Krishna Chaitnya; Macharla Madhavi; Dasari Swetha; Udatha Srinivas; N. Naga Swathi; Duggineni Indu; A. Veera Navya Vani; G. P. V. Siva Sankar; B. Saranya; and Bapatla Engineering College

MUMBAI, India, April 17 -- Intellectual Property India has published a patent application (202641043486 A) filed by Tatikonda Krishna Chaitnya Macharla Madhavi; Dasari Swetha; Udatha Srinivas; N. Naga Swathi; Duggineni Indu; A. Veera Navya Vani; G. P. V. Siva Sankar; B. Saranya; and Bapatla Engineering College, Chirala, Andhra Pradesh, on April 6, for 'telecom customer churn prediction and analysis.'

Inventor(s) include Macharla Madhavi; Tatikonda Krishna Chaitanya; Dasari Swetha; Udatha Srinivas; N. Naga Swathi; Duggineni Indu; A. Veera Navya Vani; G. P. V. Siva Sankar; and B. Saranya.

The application for the patent was published on April 17, under issue no. 16/2026.

According to the abstract released by the Intellectual Property India: "Customer churn prediction has significant importance in telecom sector. Business costs associated with retaining customers are lower than those associated with acquiring new customers. Machine learning models were utilized to perform the prediction of customer churn. The data set had 7043 samples and 21 columns, which were collected from Kaggle. In this analysis, the accuracy will not be 100 percent at any cost because it depends on human nature; there may be a possibility of retaining the customer at any time. After training and testing of the data split is over, the metrics of classification are determined. The model has been evaluated with metrics such as accuracy, precision, F1 score, and F1 measure. Machine learning models used to slove this problem were logistic regression (75%), SVM (74%), Random Forest (76%), KNN (76%), XG Boost (79%), Gradient Boost (80%). The best model in this customer churn was Gradient Boost. Along with the classification metrices, figured out the top 10 factors affecting for the churn. After obtaining the results, a graphical user interface was created, which takes the input of any sample and obtains the output results directly for the user, and if customer wants a detailed report for the churning, is there to download. This work mainly helps in telecom analysis for businesses in day-to-day life. This work helps the telecom sector to better understand about the customers and provide good services as per their requirements to avoid churn."

Disclaimer: Curated by HT Syndication.