Analyzing and Rewarding Credit Card Spending Habits in India: a Machine Learning Approach

Abstract

With the rapid adoption of digital payments in India, credit card companies are focusing on customer loyalty and planning rewards to incentivize spending, especially during peak periods like festivals. However, there is a gap in developing a tailored system that optimizes sales and reward structures for these companies. The proposed work addresses this gap by leveraging machine learning techniques to analyze and assess credit card spending patterns and propose design targeted reward programs. Besides this, this study focuses on categories as luxury, travel, groceries, EMIs payments, and others and employs ML methods, using K-Means clustering to segment users based on card types (Silver, Gold, Platinum, and Signature). Feature engineering is another key in improving the model’s understanding and providing insights, particularly in calculating reward points based on various attributes and spending behavior. The usage of original and synthetic datasets ensured scalability and adaptability across different financial domains as well. The results highlight the potential and need of ML to optimize reward allocation and provide real-time predictions, enabling financial institutions to tailor their offerings for increased customer engagement and retention. By aligning rewards with high-margin spending categories and leveraging adaptive frameworks, this study offers strategies to enhance credit card reward programs. The proposed ML model achieved an R² value of 0.99, demonstrating superior accuracy in optimizing reward point distribution.