• Version
• Download 33
• File Size 340.52 KB
• File Count 1
• Create Date 28 May 2025
• Last Updated 28 May 2025

PREDICTIVE ANALYTICS FOR TAX EVASION DETECTION IN INDIA

Authors:

Sakshi Mahesh Bobade
Student, P. E. S. Modern College of Engineering, Pune-5

Ganesh Murlidhar Belokar
Student, P. E. S. Modern College of Engineering, Pune-5

Dr. Mrs. Shivani Budhkar
Professor, P.E.S. Modern College of Engineering, Pune-5

ABSTRACT

This research explores the application of predictive analytics to enhance tax evasion detection in India, addressing a critical challenge in modern tax administration. The study develops a robust machine learning framework that integrates both supervised classification and unsupervised anomaly detection methods to analyse diverse taxpayer data—including historical tax returns, audit outcomes, and third-party financial records. Traditional methods, predominantly manual and rule-based, have proven insufficient in capturing the complexity of evolving evasion strategies. By leveraging advanced algorithms such as logistic regression, decision trees, random forests, and neural networks, alongside anomaly detection techniques, the proposed model is designed to identify subtle discrepancies in taxpayer behaviour that may indicate fraudulent activity.