Prediction method and system of outlet goods quantity, and storage medium

Abstract

The invention provides a method for predicting the amount of goods at a website and a storage medium. The method comprises the following steps: performing signal processing on cargo quantity data in apreset historical time period to obtain a fluctuation period, and predicting the cargo quantity of a next moment period by referring to fluctuation parameters of a previous period when the instantaneous cargo quantity of the next moment is predicted; and the method can adjust the fluctuation characteristic parameters according to the real-time obtained goods amount, reduce the prediction error, monitor the change of the arrival goods amount in time, and facilitate the optimal allocation of resources. The method not only can be accurate to prediction every day and per hour, but also can automatically adjust parameters according to market changes, and improves the accuracy of prediction results. In the aspect of calculation, only historical cargo volume data is needed, the data dimension issimple, the calculation complexity is low, calculation of related parameters is convenient and flexible, development and deployment are convenient, the stability is good, and the calculation speed ishigh.

Description

Technical field [0001] The present invention relates to the field of e-commerce, in particular to a method and system for predicting the volume of goods at an outlet, and a storage medium. Background technique [0002] The prior art express delivery volume forecasting methods generally use time series models and gray theory models as linear algorithm models based on statistical theory, which require high sample requirements and poor non-linear data prediction results. However, it is difficult to determine the structure of neural network, and it is easy to fall into the local extreme value. It is difficult to determine the kernel function and parameters of the support vector machine. [0003] The above methods also have low accuracy, high computational complexity, and are not accurate to the hourly forecast. Related model parameters need to be established in advance, and model parameters cannot be automatically adjusted in time, such as temporary e-commerce promotions, holidays and...

AI Technical Summary

Problems solved by technology

[0004] In view of this, it is necessary to provide a method, system, and storage medium for forecasting the volume of outlets for the problems of low precision, complex calculations, and inaccurate forecasts of traditional forecasting methods.

Images