统计信息下在线反向拍卖的竞争分析

对于投标具有统计特征的在线反向拍卖问题,利用在线算法与平均情形竞争分析相结合的方法,讨论了单一定价策略的平均情形最优单一定价及其竞争性能,提出了无限可分商品在线反向拍卖的平均情形竞争分析策略,基于此策略建立了具有均匀分布特征的在线反向拍卖模型,通过对模型求解得到了采购商的竞争需要曲线。与不考虑投标的统计信息、只是利用常规的最坏情形竞争分析得到的在线反向拍卖的竞争策略进行对比分析,发现统计信息的利用提高了在线反向拍卖策略的竞争性能。

Competitive Analysis of Online Reverse Auctions with Statistic Information

For online reverse auctions with probability distribution bids from sellers, using online algorithm and average-case competitive analysis, we discuss the average-case optimal single pricing and the competitive performance of single price strategy, and then propose average-case competitive analysis strategy of online reverse auctions for unlimited divisible goods. We build a model of online reverse auctions on the base of the strategy, and obtain the competitive demand curve of the buyer through solving the model. Moreover, by comparing the average-case analysis strategies with the conventional worst-case competitive analysis which ignore statistic information of bids, we conclude that the competitive performance of the strategies is improved.