An Improved Hybrid Model base on SVM and Random Forest for the Prediction of Corporate Taxation

With the development of the economy and society, the level of the tax burden of enterprises has become one of the important influencing factors for the profit level and future development trend of enterprises. Therefore, how to reasonably carry out corporate tax planning has become a hot issue in the industry as well as in academia. In this study, a new model combining support vector machine and random forest model is proposed for predicting the future tax amount of enterprises. The experiment uses a public dataset of an enterprise with 19 fields and each field contains 120 data. The dataset is first statistically analyzed and preprocessed. Secondly, the dataset is predicted using three models: support vector machine, random forest, and new model, respectively. Finally, the performance of each model is evaluated using ME, RSME,MAE. This study can improve the efficiency of enterprise tax forecasting, reduce the waste of resources in enterprise tax forecasting, and provide some insights for enterprise tax planning.