Pricing analysis in online auctions using clustering and regression tree approach

Publication Type:
Conference Proceeding
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2012, 7103 LNAI pp. 248 - 257
Issue Date:
Full metadata record
Auctions can be characterized by distinct nature of their feature space. This feature space may include opening price, closing price, average bid rate, bid history, seller and buyer reputation, number of bids and many more. In this paper, a price forecasting agent (PFA) is proposed using data mining techniques to forecast the end-price of an online auction for autonomous agent based system. In the proposed model, the input auction space is partitioned into groups of similar auctions by k-means clustering algorithm. The recurrent problem of finding the value of k in k-means algorithm is solved by employing elbow method using one way analysis of variance (ANOVA). Based on the transformed data after clustering, bid selector nominates the cluster for the current auction whose price is to be forecasted. Regression trees are employed to predict the end-price and designing the optimal bidding strategies for the current auction. Our results show the improvements in the end price prediction using clustering and regression tree approach.
Please use this identifier to cite or link to this item: