Data Mining in Financial Analytics to Forecast Loan Behavior: An Integrated Approach Combining Time Series and Machine Learning

Abstract

This study aims to apply time series as a predictive data mining technique to analyze future loan trends based on historical customer behavior. Using the Power BI business intelligence tool, the research presents a logical analysis of the financial status of the selected public bank. It also seeks to support the enhancement of decision-making quality and direction. The study concluded that state-subsidized loans are expected to rise significantly. In contrast, loans requiring personal contributions are projected to remain relatively stable. Partially subsidized loans, however, are likely to fluctuate in response to economic conditions and financial policies.