Research And Improvement of E-Commerce Shipment Prediction Model Based on Neural Network

Abstract

In this paper, for the problem of e-commerce shipment prediction, the first initial modeling and solving is based on BP neural network model. Through the construction and training of the neural network model, the relationship between the commodity shipments of each merchant in different warehouses and various properties is explored and predicted. Then, the irrationality of the initial model is improved, and the STL-LSTM deep learning algorithm is used for prediction model building to better capture the seasonal characteristics and trends of the time series data and improve the prediction accuracy. Finally, combined with the SARIMA neural network prediction model, the e-commerce shipments are predicted and analyzed more accurately to provide a scientific basis and reference for the merchants' demand for goods in different warehouses. Through the research and improvement of this paper, the accuracy and practicality of e-commerce shipment prediction can be effectively improved, which has certain theoretical and practical significance.