Business volume prediction model construction method, and business volume prediction method

Abstract

The invention relates to a business volume prediction model construction method and device, computer equipment, a storage medium, and a business volume prediction method. The method comprises the following steps: acquiring historical business volume data; constructing a historical sample set based on the historical business volume data; performing discrete processing and complete up-sampling processing on the historical sample set to generate a training sample set; training a preset initial random neural network through the training sample set and a dynamic sampling algorithm, and constructing a business volume prediction model. According to the invention, the business volume prediction model is constructed through dynamic sampling and the random neural network, and the dynamic sampling balances the training samples in a peak period and a non-peak period according to a prediction error in the model training process, so that the imbalance problem of too few samples in the peak period is solved, the obtained business volume prediction model can effectively solve the problem of inaccurate business volume prediction in the peak period, and a more accurate basis for decision making in the peak period is provided.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a traffic forecasting model building method, device, computer equipment and storage media, and a traffic forecasting method. Background technique [0002] With the rapid development of computer technology and e-commerce, machine models for traffic forecasting are also constantly evolving. For example, in the field of logistics, the business volume of daily shipment volume is an important reference for decision-making points such as branch scheduling and vehicle scheduling. Therefore, the accurate prediction of the future daily shipment volume is of great significance for the stable operation of all links in logistics distribution. [0003] The fbProphet algorithm is widely used in business volume forecasting, for example, it can be used to predict the daily shipment volume of the logistics industry. However, the algorithm has higher accuracy for traffic forecasting ...

AI Technical Summary

Problems solved by technology

[0004] Based on this, it is necessary to address the above technical problems and provide a traffic forecasting model construction method, device, computer equipment, storage medium and traffic forecasting method that can solve the problem of low traffic forecasting accuracy of existing models for peak periods

Images