PEER TO PEER LENDING LOAN GRADE AND INTEREST RATE PREDICTION USING APPLICATION FINANCIAL SIGNALS

Abstract

Peer-to-peer (P2P) lending has emerged as an innovative financial technology that
directly connects borrowers and lenders through online platforms, eliminating the
need for traditional intermediaries such as banks. While these platforms provide easier
access to loans and attractive investment opportunities, accurately assessing borrower
credit risk remains a significant challenge. Determining appropriate loan grades and
interest rates is crucial to minimize financial losses, ensure platform reliability, and
build trust among users.
This project proposes a machine learning-based approach to predict loan grade and
interest rate using borrower application data and financial indicators. The system
utilizes key financial attributes such as annual income, employment length, debt-toincome
ratio, credit history, loan amount, and loan purpose to evaluate borrower risk.
Data preprocessing techniques—including handling missing values, normalization,
and encoding of categorical variables—are applied to prepare the dataset. Feature
selection is then performed to identify the most relevant factors influencing loan
grading and pricing.
Machine learning algorithms are trained on historical lending data to capture hidden
patterns and relationships between financial signals and loan risk levels. The trained
model can accurately classify loan grades and estimate suitable interest rates for new
loan applications. Performance is evaluated using appropriate metrics to ensure
reliability and effectiveness.