Research on the Value Retention Rate Evaluation of Pure Electric Used Cars Based on Deep Learning Models

This study proposes a deep learning model-based method for evaluating the residual value of pure electric vehicles, addressing issues such as low residual value and difficult valuation. Based on the big data of pure electric second-hand car transactions, evaluation models are constructed using three algorithms: TabTransformer, LGBM, and SVR. TabTransformer model is used to evaluate the value retention rate of pure electric vehicles, providing reference for second-hand car dealers, financial institutions, and consumers.