The D-S proof theory is used to incorporate six different advantages of the basic classifier. There are many studies on machine learning and data mining algorithms to improve the impact of credit risk assessment. However, there are few methods that can fulfill its universal and effective properties. This paper proposes a new model for multi-classification assessment of personal credit risk based on information fusion theory (MIFCA) using six machine learning algorithms. The MIFCA model can simultaneously integrate the advantages of multiple classifiers and reduce interference from uncertain information. To verify the MIFCA model, a set of data collected from a set of actual data from a commercial bank in China. The experimental results show that the MIFCA model has two remarkable points in different evaluation criteria. One is that it has greater accuracy for multi-classification assessment, and the other is that it is suitable for different risk assessments and is universally applicable. In addition, the results of this study can also provide benchmarks for banks and other financial institutions to strengthen their risk prevention and control capacities, improve their credit risk identification capabilities and avoid financial losses. The new model has greater accuracy and universality than other classifiers. A new multi-classification model for loans held by banks is proposed. .