Research on Enterprise Financing Risk Assessment by Pruned Decision Tree Algorithm

From the past few years, enterprise financing risk assessment involves in identifying, evaluating and managing potential risks which impacts a company’s investments and overall stability. Traditional approaches for financing risk assessment had faced challenges like focused on specific assets and failed to adequately identify and manage risks across the entire organization. Therefore, this research proposes pruned Decision Tree (DT) algorithm for enterprise financing risk assessment. Initially, the data is collected from Lending Club Loan dataset which contains a detailed information on business loans. Then, this data is fed into classification with DT to classify the businesses into risk categories by analyzing financial attributes, optimizing decisions through entropy-based splits. Next, pruning is introduced to remove unnecessary branches and nodes where it prevents overfitting. After that, the risk assessment system is established based on multi-level comprehensive evaluation which evaluates multiple risk indicators through Analytic Hierarchy Process (AHP) and fuzzy matrices. The proposed Pruned DT attained better results in terms of accuracy (93.0%) when analyzed with existing DT.