[2H1-J-2] Machine learning: modeling in businesses

[2H1-J-2-03] An Estimation Model of Exhibits Price for Second-hand Fashion Items Based on Sales History Data

〇Izumi Kuwata¹, Tomoya Sugisaki¹, Kenta Mikawa², Masayuki Goto¹ (1. Waseda University, 2. Shonan Institute of Technology)

Keywords:The second-hand fashion EC site, Estimation of Exhibits Price, Random Forests, Sales History Data

In this study, we focus on the second-hand fashion EC site. In this site, the trade-in price is determined according to the exhibition price so that the decision of the appropriate exhibition price is strongly important. However, it is difficult to determine the best exhibition price because there are two types of items such as the items purchased with no discount and that purchased with discount. Therefore, we proposed a method to estimate the maximum exhibition price without discount for each item. In our proposed method, at first, we construct a binary classifier which classifies the item which is purchased with a discount price or not. In addition, we estimate the best price with changing the exhibition price of each item little by little. The effectiveness of our proposed method is clarified using real purchasing history data of this site.