【作者】 李成堃；

【导师】 张勤生；

【作者基本信息】 中国海洋大学 ， 管理科学与工程， 2009， 硕士

【摘要】 网络拍卖是传统拍卖借助互联网发展起来的有效资源配置方式与价格发现机制。多属性拍卖是一种在拍卖中考虑买卖方的多重属性的网络拍卖,可以有效弥补传统拍卖仅就价格属性协商交易的不足;随着网络拍卖的发展,双向拍卖是拍卖发展的大势所趋,因此,结合两种拍卖形式的双向多属性拍卖将是网络拍卖以后研究的重点。本文尝试设计双向多属性拍卖的最优机制,通过建立数学模型求解双向多属性拍卖的最优机制,并将双向多属性拍卖的最优机制从一次交易中买卖方只有单位拍卖品交易扩展到团购交易中来。论文的研究工作主要包括:建立双向多属性拍卖的一般模型,提出二维拍卖模型的设想,将多属性拍卖的多维模型简化成双向二维拍卖模型,并提出双向多属性拍卖机制设计的一般思路。研究机制设计控制与协调策略问题,利用神经网络将买卖方的非价格属性集转化为一个质量指标,从而将多属性拍卖模型简化为二维模型,根据量化得到的买卖方质量指标来确定进入到拍卖中来的竞胜标,满足一定质量指标的买卖方进入到拍卖中,以此实现拍卖机制的筛选机制。然后研究在卖方一次交易中交易多个拍卖品的情况,团购的产生以及团购对拍卖机制设计中的分配和支付机制的影响。通过上述研究工作,获得了以下研究成果与研究结论:通过建立双向多属性拍卖的线性规划模型,结合激励相容理论、个体理性等机制设计理论,得到双向多属性拍卖的最优机制,包括筛选机制、报价机制、分配机制以及支付机制。筛选机制通过人工神经网络实现,利用人工神经网络量化买卖方质量指标,并根据质量指标确定进入到拍卖中的买卖方。根据个体理性分析买卖方的报价机制,即买卖方的报价属于一个合理的报价区间,买卖方的报价满足个体理性约束。最后建立双向多属性拍卖的数学模型将最优拍卖机制设计问题转化为约束条件下的目标规划求最优解的问题。求解此数学模型,得到最优拍卖机制下的分配机制和支付机制,然后,利用一个简单的拍卖算例对最优拍卖机制的市场效率进行了验证,同时对团购模式下的拍卖市场效率分情况进行了求解。更多还原

【Abstract】 As an effective mechanism to allocate resource and discover price, online auction has already grown up with the development of internet. Multi-attribute auction is a form of online auctions which considers multi-attributes of the buyers and the sellers, multi-attribute auction can make up for a deficiency of traditional auction which consult on price property only, Two-way auction is the development trend of online auction with the development of the internet, Therefore, two-way multi-attribute auction must be the focus of the future study.So I try to design an optimal mechanism for two-way multi-attribute auction by using a mathematic model and apply the optimal mechanism from the universal model to group buying model.The main content of my thesis are carried out from the following aspects: set up a general model of two-way multi-attribute auction, then propose the idea of two-dimensional model of two-way multi-attribute auction, simplify the multi-dimensional model to two-dimensional model, and post the main framework of optimal mechanism. Then, study how to control mechanism and coordination strategy. Using neural networks to transform non-price attributes to a quality property, which will simplify the multi-dimensional auction model to two-dimensional model. Screening mechanism regulate that only the buyer (or seller) who gets a higher quality property can get into the auction. At last, discuss the influence to optimal mechanism of group buying when a seller is selling a lot of goods in one transaction.Through the research mentioned above, the following conclusions has been achieved: Get the optimal mechanism of two-way multi-attribute auction through setting up a mathematic model and combining the theory of incentive compatibility and individual rationality, including the screening mechanism, quotation mechanism, distribution mechanism and payment mechanism. Screening mechanism is realized by using artificial neural network; analysis the quotation mechanism of auction participant according to the theory of individual rationality, the quoted price must be rational and belongs to a certain region; payment mechanism and distribution mechanism are achieved by solving the mathematical model which change the optimal auction mechanism designing problem into a goal programming under the conditions of the planning for the optimal solution of the problem. Then, use a simple example to verify the optimal mechanism of two-way multi-attribute auction, and calculate the market efficiency of group buying auction market.更多还原