Research on used car valuation problem based on model fusion

In recent years, with the rapid development of the automobile industry, the trading volume of second-hand cars in our country has grown rapidly. However, with the continuous expansion of the second-hand car market, a scientific and reasonable evaluation system or unified standard has not yet been formed in the second-hand car market, which makes the second-hand car trading market lack credibility and restricts its development of the second-hand car trading market. Therefore, it is particularly important to establish a reasonable and perfect second-hand car valuation method. In this paper, GBDT, LightGBM, and XGBoost models are introduced into the field of the used car valuation, and by analyzing the influence of body infrastructure and vehicle conditions, a used car valuation model based on the fusion of GBDT, LightGBM, and XGBoost is constructed. Then it conducts in-depth analysis and research on the problem of used car valuation. At the same time, to verify the advantages and rationality of the model proposed in this paper, the used car valuation model based on the fusion of GBDT, LightGBM and XGBoost is compared and analyzed with random forest, KNN, linear regression, and other models. Finally, after verification, the proposed model based on GBDT, LightGBM, and XGBoost fusion can significantly improve the prediction accuracy, and under the self-defined model evaluation standard in this paper, the model recognition accuracy is up to 89%, which has good practical value.