New Energy Vehicle Sales Prediction Based on Data Mining: A Case Study of BYD

With the rapid development and transformation of the vehicle industry, the production intensity of vehicle manufacturers is increasing year by year. Blind production planning and increasingly fierce competition between enterprises make enterprises prone to the problem of unbalanced production and sales. Due to the development of the Internet, customers tend to obtain and share product information through online media before and after purchasing products, which provides a new idea for the sales prediction of new energy vehicles. Based on this, this paper uses text mining and neural network method to construct a multi-feature time series prediction model. Taking BYD enterprise as an example, this paper provides decision-making reference for the production and marketing plan of new energy enterprises. It is found that the Attention-Seq2Seq model has higher prediction accuracy than GRU network and Seq2Seq model. This paper proposes a new energy vehicle sales prediction model based on data mining, which helps enterprises to identify changes in new energy vehicle sales in time and optimize the inventory of production enterprises.