Artificial Intelligence Credit Risk Assessment Model Based on MLP‐Hybrid Clustering

For the credit risk problem, an artificial intelligence credit risk assessment model based on multilayer perceptron (MLP) hybrid clustering is proposed. This study is the first to evaluate credit risk evaluation indexes based on a combination of natural language processing (NLP), cluster analysis, and correlation analysis and to construct a credit risk evaluation index system. The weights of the indicators are then determined through the application of the analytic hierarchy process (AHP) and the criteria importance through intercriteria correlation (CRITIC) assignment method; then, the weighted credit evaluation indicator data are used to construct a credit risk assessment model based on the MLP to predict the credit status of the business entity, and then a Gaussian mixture model (GMM) combined with the expectation–maximization (EM) algorithm is employed to divide the credit risk levels into intervals. Finally, based on the research information of 246 family farms and ranches in 12 cities in the Inner Mongolia region, an empirical analysis is conducted. The results demonstrate that (1) most of the investigated samples of family farms in Inner Mongolia are in a low‐risk state, which follows the law of large numbers and practical significance. (2) The accuracy of the regression model to predict the credit risk of family farms in Inner Mongolia is examined with the ROC curve, and the results show that the AUC value is 0.92, and the MLP model fits well.