A Method for Predicting Monthly Sales of New Energy Vehicles Based on Weighted Ensemble Learning
The accurate prediction of car sales and analysis of influencing factors have a certain reference significance for the indus-trial layout of automotive enterprises and the purchase of consumers.This article crawls typical new energy vehicle configuration data,evaluation data,and sales data from the past five years,and constructs a training dataset containing 127 car companies and 1 440 car models based on data cleaning and feature engineering methods.Then,an ensemble prediction model for new energy vehicle sales,combining SVR algorithm and LightGBM algorithm using an weighting strategy,is designed to accurately predict the monthly sales of new energy vehicles for each car company.The sales prediction model achieves an average R2 value of 0.92 on the total monthly sales for each car company.Experimental results demonstrate that the method proposed in this article can effectively predict the sales of new energy vehicles,providing reference information for automotive enterprises and consumers.
NETL
NSTL
万方数据
