Financial Risk Management using Machine Learning Method

AI in banking industries could solve the dilemma, including fraud, credit crisis, and financial risk management, faced by the recent market structure. In addition, AI could reduce the rates of errors committed by humans and enhance the total transparency of each dealing, replacing the routine and monotonous tasks performed by humans in the early stage of financial activities. In this study, both unsupervised and supervised learning, including SVM, logistic regression, random forest, and K-means, are performed for credit rating and loan decision issues. Through the experimental result, we conclude that the SVM algorithm, whose accuracy is up to 78.4%, has the best performance among these supervised learning models. Moreover, the three clustering results produced by K-means are more desired and provide a constructive suggestion to deal with SME credit rating. This study gives an AI-based strategy for loan decisions and credit rating, which can help bank industries solve financial issues.