The activities of online auction produces a large number of transaction data. If utilized properly, these data can be of great benefit to sellers, buyers and site administrator. Typically, prediction result may help sellers optimize the selling price of their items. Thus, transaction time can be shorted and cost can be saved. This paper collects a lot of historical exchange data from Eachnet, an online auction website most famous in China, and uses machine learning algorithms to forecast the final price of auction items. On the basis of the categorization and preprocessing of data, the prediction is made. The prediction results and performances are discussed to verify the proposed solutions.
[1]
J. Goeree,et al.
Winner's Curse without Overbidding
,
2003
.
[2]
Thore Graepel,et al.
Invariant Pattern Recognition by Semi-Definite Programming Machines
,
2003,
NIPS.
[3]
Alexei Pozdnoukhov,et al.
Tangent vector kernels for invariant image classification with SVMs
,
2004,
ICPR 2004.
[4]
Bernhard Schölkopf,et al.
Incorporating Invariances in Support Vector Learning Machines
,
1996,
ICANN.
[5]
Wolfgang Jank,et al.
Profiling Price Dynamics in Online Auctions Using Curve Clustering
,
2005
.