Credit risk is the risk that arises because the borrower does not pay all or part of the receivables or does not pay on time and will cause losses to the company.
The distribution of loans with problem loans will affect liquidity. Due to the emergence of non-performing loans, cash that should have entered and added to the company's liquidity did not occur, resulting in the company's being unable to meet its short-term obligations.
To minimize this credit risk, a process called credit scoring and credit rating is usually carried out on the borrower. The output of this process will be the basis for determining whether a new loan application is accepted or rejected.
The process of obtaining the output is carried out by using a machine learning method approach, namely a decision tree that focuses on classification.
The conclusions obtained from this analysis are:
- 76% of the training data was successfully classified into the correct risk rating class, while 24% of the other data still had errors in class placement.
- Based on 100 pieces of training data, the machine learning model that was built succeeded in classifying 83 pieces of data with the right risk rating.
So based on these conclusions, it can be ascertained that the risk rating prediction results from the machine learning model built are expected to help speed up the process of determining whether new loan applications are accepted or rejected.