Research on Cargo Volume Prediction Based on Prophet-RF Model

The Prophet prediction model excels at working with data with distinct cyclical and seasonal characteristics, while the Random Forest model excels in predicting random fluctuations in data due to its flexibility and strong ability to identify outliers. Therefore, this paper aims to combine the advantages of these two models and explore the combined application methods of the two models, taking the background of cargo volume in the sorting center, which is both strongly cyclical and seasonal, accompanied by random fluctuations, as an example. Firstly, the time series prediction model based on Prophet and the residual prediction model based on random forest were constructed respectively, and their prediction accuracy was verified. Then, the Prophet-RF combination model is used to integrate the two models to obtain more accurate and reliable predictions. This research method not only provides a scientific tool for the prediction of cargo volume and resource allocation of sorting centers, but also provides a useful reference for the prediction and application of similar scenarios such as transportation and e-commerce.