Sales prediction method by least squares support vector regression

摘要

Sales prediction method based on least squares support vector regression is presented to predict the sales in the paper.In the study.mechanical production sales are used to study the prediction ability for sales by least squares support vector regression.Least squares support vector regression prediction model includes 3 input nodes and 1 input node.We employ MAPE to evaluate the least squares support vector regression prediction model.In order to.BP neural network is applied to predict the sales to show the prediction performance of the least squares support vector regression prediction model compared with BP neural network.The experimental results indicate that least squares support vector regression is very higher accuracy than BP neural network.