Second-hand vehicle value estimation method based on smooth support vector regression fusion

Abstract

The invention discloses a second-hand vehicle value estimation method based on smooth support vector regression fusion, and the method comprises the following steps: obtaining a transaction price during the transaction of a second-hand vehicle, taking the transaction price as a data source, and carrying out the processing, cleaning and desensitization of the data source; taking transaction pricesin the desensitized data sources as input vectors x, classifying the transaction prices according to different influence factors, taking the influence factors as relative variables y, and forming a set s{xi, yi}; performing constraint optimization on the set; and applying a machine learning technology and smooth support vector regression and performance fusion to obtain a fusion prediction model,and performing analysis by adopting the model to obtain a predicted transaction price of the second-hand vehicle. A set of data sources is constrained and optimized by utilizing a smooth support vector regression algorithm, and a machine learning technology and the smooth support vector regression algorithm are fused, so that the technical problem that the valuation of a second-hand vehicle is notaccurate enough due to the fact that prediction is based on increasingly old statistical parameter values after the constraint optimization of the data sources can be solved.

Description

technical field [0001] The invention relates to a used car valuation method, in particular to a used car valuation method based on smooth support vector regression fusion. Background technique [0002] In recent years, the volume of second-hand car transactions in my country has expanded rapidly. In 2000, the transaction volume was only 250,000 vehicles, and by 2018 it had reached 13.82 million vehicles. The large middle and bottom income earners are an important consumer group of used cars. While the scale of the second-hand car market continues to expand, the price evaluation system of my country's second-hand car market has exposed the problem of not being able to adapt to market demand. The randomness and more human factors in the price evaluation have affected the reasonable evaluation of used cars and restricted them. With the orderly development of the second-hand car market. In contrast to foreign developed countries, the sustainable, healthy and stable development ...

AI Technical Summary

Problems solved by technology

[0003] The object of the present invention is to provide a kind of used car valuation method based on smooth support vector regression fusion with faster convergence speed, better fitting accuracy and good predictive performance, to solve the inaccurate in the above-mentioned background technology Constrain the transaction price and influencing factors in the data source and solve the data source constraints. After optimization, the forecast is based on the statistical parameter values ​​that are getting older and older, which leads to the technical problem that the valuation of used cars is not accurate enough

Images