Financial Risk Warning in Listed Manufacturing Enterprises Using a Huffman Tree Enhanced Support Vector Machine with Arithmetic Optimization

Abstract

During the production and operation process, manufacturing enterprises may experience financial instability due to factors such as capital flows, cost control, and market changes, which can affect their profitability and debt-paying ability. Although certain progress has been made in financial risk early warning, there are still obvious problems such as lagging early warning and incomplete indicator systems. To further optimize the early warning mechanism of enterprise financial risks and improve the response efficiency, an improved Huffman tree support vector machine algorithm is proposed. This algorithm combines arithmetic optimization algorithms and is applied to the early warning and control of financial risks in listed manufacturing enterprises. This method converts the low-dimensional space into a high-dimensional space through nonlinear mapping, thereby enhancing the computing speed and prediction accuracy. The study adopts five publicly available multi-class imbalanced datasets. The experimental results showed that the accuracy rates of the improved Huffman tree support vector machine algorithm on the training set were 80.3649%, 89.6989%, 90.3654%, 96.2453%, and 97.4658% respectively. The accuracy rates on the test set were 85.3694%, 91.3658%, 92.3654%, 94.2652%, and 96.7659% respectively. The prediction accuracy of the overall model reached 81.8%, which was higher than that of traditional methods. The results show that the optimization algorithm combining Huffman tree mechanism and support vector machine can effectively meet the needs of financial risk early warning in manufacturing enterprises, providing theoretical support and practical basis for subsequent financial risk diagnosis and control applications.