Prediction Online Auction Price Using Functional Data Analysis

Abstract

The goal of the proposed system is to derive models for forecasting the final price of ongoing auction. The forecasting task is important not only to the participants of an auction who compete against each other for the lowest price, but also to designers of bidder side. Forecasting price in online auctions is challenging from statistical point-of-view because traditional forecasting model do not apply. The reason for this are three typical feature of online auction data: (1) unequally spaced bid; (2) the limited time of an auction; (3) the dynamic of bidding change significantly over time. The key feature of our model is that it operates during the live –auction. Auction data typically arrive as a sequence of bids over a period of time. Taking a functional data analysis approach, the system acts the bid from a single auction as recognition from a continuous price.