Comparison of Multiple Machine Learning Models Based on Enterprise Revenue Forecasting

Enterprise operating income is an important part of enterprise revenue. It is of great reference significance to realize the prediction of corporate income for corporate operating income management. However, the corporate income forecast process of most companies is time-consuming and error-prone because corporate revenue forecasts are calculated manually by hundred of financial analysts. Moreover, it is also difficult to forecast through traditional statistical methods because of the data noise that usually exists in such data, as well as the high dimensionality of the data. At the same time, the data set used in this paper is relatively small, so models such as neural networks with more strict data volume requirements are not suitable. To address the above problems, this paper proposes to use multiple models such as support vector machines to predict the business revenue data of enterprises on the basis of relatively controllable model complexity. And then, the prediction ability of the models is evaluated relatively reasonably using three indexes, mean absolute error (MAE), root mean square error (RMSE), and absolute error value (MAPE).