This article analyzes current issues in bank credit risk assessment and focuses on the scoring model created using data analysis methods to assess borrower creditworthiness. Credit scoring is chosen as an example to illustrate the data processing process. Credit risk assessment has become an important tool for financial institutions to distinguish between customers who can or cannot repay their loans. This is why machine learning algorithms are successfully used in credit scoring. Credit risk mitigation is an area of ​​growing interest since the financial crisis, with financial institutions collecting large amounts of data. Therefore, risk analysts are faced with the difficult task of processing large amounts of data and adequately determining individual creditworthiness. Financial institutions can use advanced machine learning techniques to develop highly efficient credit scoring models. This paper uses machine learning classification techniques for benchmarking. The results show that regression is the initial best guess, followed by the random forest model. The widely used logit model performed better than vector machines. We also show that other machine learning methods outperform the widely used logit model in terms of classification ability using the Kolmogorov-Smirnov test.