The rapid growth of online lending industry has led to many problems due to the uneven quality of enterprises and credit risks. These problems have exposed the insufficient supervision of the relevant enterprises and the insufficiency of their risk control ability. Therefore, this paper takes online lending enterprises as an example, constructs the assessment index system from the four aspects of business status, capital security, basic information, and platform credit enhancement, and utilizes the rotating forest algorithm to carry out feature transformation to increase the influence of features. To add feature space randomness and learners’ diversity, the transformed feature space is used to train the base learner of decision tree algorithm, support vector machine algorithm, logistic regression and neural network algorithm, etc. Then XGBoost is used as the secondary learner for Stacking integration. The experimental results show that the model proposed in this paper has good prediction and generalization ability.