The standardized strategy of the enterprise economic management model is related to the enterprise’s success or failure in the market competition, but it is difficult to detect the problems in time when they occur, and the development of the performance decline and heavy debt when the accumulation of weight is difficult to return. This paper proposes a risk analysis model based on machine learning for enterprise management, which is first based on SVM algorithm using principal component analysis for dimensionality reduction, and then introduces the Gaussian kernel function to improve the classification accuracy, and combines it with the decision tree C4.5 algorithm to complete the construction of enterprise economic management risk analysis model. Next, the enterprise economic management risk evaluation index system is constructed, and the assignment operation is carried out using the entropy weight method. On this basis, after testing the performance of the model using publicly available corporate financial disclosure data, Company A is taken as an example for the analysis of the enterprise management model. It has been found that the comprehensive evaluation scores of the sample companies in the model from 2021 to 2023 are 72.35, 68.393, and 67.011, all of which belong to the risk-free level. However, from 2021, the sample company’s gearing ratio and cash current liability ratio are 4.398 and 2.508, with low scores, and the company’s debt pressure begins to show, while from 2021 to 2023, the earned interest multiplier decreases from 14.18 to 9.019 points down by more than 5 points, which indicates that the pressure to pay interest is gradually having a negative impact on the sample company. This study provides a new method for companies to improve their economic management model and strategy, which is conducive to analyzing business problems and enhancing competitiveness.