Recommended Analysis Of Second-hand Commodity Price Of Electronic Mall Based On LightGBM

The use of electronic channels for the sale and purchase of goods has become popular today,and more and more young people are choosing online shopping as their preferred shopping method.At the same time,more consumers have begun to use the online sales platform to sell personal goods(second-hand goods).Limited by the professionalism of sales and the integrity of the information obtained,the pricing strategy of individual merchants when selling second-hand goods often lacks rationality.Based on the lightGBM model,this paper attempts to give the recommended price of a new personal merchant when selling its second-hand goods from the historical transaction price of second-hand goods.In this paper,lightGBM model is used to analyze the data of Mercari,a second-hand trading platform,so as to provide guidance and suggestions for second-hand merchants to release reasonable commodity prices.Due to the data of Mercari second-hand platform is open and transparent,data of Mercari second-hand platform is taken as the research object.LightGBM algorithm is a fast,distributed,high-performance decision tree algorithm based on the idea of histogram gradient lifting framework.It has a higher predictive ability in classification and regression.In this paper,data collection,preprocessing,model parameter design,model evaluation,second-hand commodity price forecast and other aspects will be carried out.In the empirical analysis part,after data separation and transcoding of the collected data,the paper makes a simple descriptive statistics on the transaction data,including the summary of the use traces of goods,the summary of trade names,freight payers and price distribution.The lightGBM model was built after the descriptive statistical analysis of the classification attributes of commodities.The results show that the lightGBM model performs well and accurately predicts the price of second-hand goods.This paper innovatively applies the lightGBM model based on decision tree algorithm to analyze the data of second-hand shopping malls and further predicts the price of second-hand goods with given information.Applying the model to a large number of data in the mature second-hand market will provide guidance and suggestions for the subsequent research.