The principle of the model is to track the bid record and sales record of each used car through search engine technology, determine the bid weight of the merchants through the successful transaction volume of the used car merchants, and then make weighted average and linear regression analysis on the bids of a large number of used car merchants. So as to get the approximate price of specific models (80% is close). Finally, the "fair price" of a used car can be specifically and accurately determined by evaluating the car condition (insurance, maintenance level, specific parameter configuration, quotation of the same new car, etc.). 20%). Because the accurate estimation is completed in two stages, the revised model is also named "two-stage weighted model".
The linear weighted model of big data effectively solves the serious problem of confusion and lack of credibility in the third-party evaluation standard of domestic used car trading market. It provides a feasible valuation method for the second-hand car market in China.