Development of a data-driven decision support system for credit risk assessments in Indian Small and medium enterprises
Credit rating is performed to estimate the performance of companies in SME sector. Banks and Financial institutions use process driven protocols and documentary evidence for giving credit rating to SMEs, which is often a time-consuming and costly affair. As an alternative, strategy soft computing techniques have found preference in many financial sectors as approach of credit risk assessment in cases where time is constrained. In this study, the SME’s credit rating data was collected from CMIE Prowess database and Ace Analyzer web portal between the financial years 2013 to 2021. The dataset was initially subjected for pre-processing to remove the outliers. Further various classification and clustering techniques were tested for predicting the nature of credit risk. On the basis of accuracy, the top two ranked clustering algorithms (i.e. k-means and hierarchical) were deployed as a predictive tool for the credit rating of SME dataset in R programming language and python.