Optimization method of internal financial budget execution and control based on SVM and clustering algorithm

With the rapid growth of enterprise financial data, traditional budget execution management methods that rely on manual analysis are unable to identify complex financial risks in a timely manner. Moreover, existing intelligent financial analysis methods mainly focus on data mining and lack comprehensive evaluation capabilities for departmental budget execution quality. To address these issues, this study develops an internal financial budget execution and control optimization system based on support vector machine (SVM) and K-means clustering algorithms. The system uses the improved K-means based on Fireworks Algorithm (FWA) to mine the financial data, and obtains the characteristics and risks of enterprise financial execution. Then it establishes and improves the department final account evaluation system, uses the analytic hierarchy process (AHP) to determine the weight of various indicators, and uses SVM to conduct scientific and comprehensive department final account grade evaluation. Based on the evaluation results, the proposed method can provide a reference for optimizing internal financial budget execution and control. On the current dataset, FWA-K-means achieved the highest mean silhouette coefficient of 0.67 ± 0.01 when k = 3. Under the original 70%/30% data split, the SVM model achieved a test accuracy of 93.67%, while 5-fold cross-validation yielded a mean accuracy of 94.00% ± 4.18% (95% CI: 88.81%-99.19%), indicating relatively stable evaluation performance on the current sample set.