Personal Finance Tracker with AI-Driven Savings Prediction
Dr. Sunil Khullar
Department of Computer Science and Engineering, KCC
ITM, Greater Noida, UP
sunilkhullar222@yahoo.co.in
Raushan Raj
Department of Computer Science and Engineering, KCC ITM,
Greater Noida, UP
raushan.rajsk04@gmail.com
Dishant Seth
Department of Computer Science and Engineering, KCC ITM,
Greater Noida, UP
sethdiishant0502@gmail.com
Abhishek Oli
Department of Computer Science and Engineering, KCC ITM,
Greater Noida, UP
Abhishekolirocks12@gmail.com
Abstract
Financial management at the individual level has been growing more difficult in the contemporary digital economy because of the prevalence of online transactions, subscription services, the lack of regular income, and the diversification of spending habits. The conventional personal finance applications are mostly recurrent expense and manual budgeting and are very shallow in their analysis and provide no foresight. Consequently, users tend not to plan their future financial results, and they end up spending more, saving less as well as poor financial planning on the long run.
In this study, the design, development, and testing of a Personal Finance Tracker, which includes AI-Driven Savings Prediction, an intelligent financial management platform that combines manual transaction tracking with machine learning-based saving prediction will be presented. The suggested system has a modular structure with transaction manager, budget management, visualization dashboards, and a specific machine learning engine. Preprocessing techniques, feature engineering, and time-series indexing are used to process historic information on the income and expenses of a business. The ensemble forecasting model is based on the use of a Random Forest regression to forecast future saving through the learning of individual user financial behaviour and temporal spending patterns.
Experimental analysis on real-world inspired data show that the predictive performance is high with the R 2 of 0.81 and low values of Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). Beside numerical forecasts, the system creates a context-driven forecasts like over spending notices, budget threshold notices, and tailored suggestions of how to enhance financial discipline. The results show that by incorporating artificial intelligence into personal finance applications, it is possible to transition to a more proactive and data-driven approach to financial planning instead of the reactive financial monitoring. This study will add to the development of smart FinTech systems by proving scalability, security, and flexibility in the context of individual financial management.
Keywords: Machine Learning, Personal Finance, Savings Prediction, FinTech, Predictive Analytics, Random Forest.
