Blockchain-Based AI Models for Credit Scoring and Risk Assessment using Fog Computing Infrastructure
Anita Kori Anita Kori, Nagaraj Gadagin · International Journal of Research Publication and Reviews · 2024
This paper presents a novel framework, Blockchain-Based Deep Learning Model LSTM-X, for enhanced credit scoring and risk assessment in financial services.Leveraging the Long Short-Term Memory (LSTM) neural network's ability to analyze time-series data, LSTM-X evaluates borrowers' creditworthiness by identifying complex, non-linear patterns within historical financial transactions.Integrating blockchain technology ensures that the credit data used for analysis is secure, transparent, and immutable, fostering trust in both data integrity and model predictions.The decentralized blockchain framework allows secure, multisource data sharing across financial institutions, improving model performance with enriched, comprehensive data.LSTM-X provides superior predictive accuracy compared to traditional models, particularly in handling irregular and noisy credit histories.This solution offers financial institutions a robust and privacy-preserving tool for assessing risk, aiding in more accurate and fair credit decisions, while adhering to regulatory compliance through verifiable, transparent data trails.