Research On Dynamic Auction Mechanism Design With Dynamic Population And Information

Auction is a general means to study many problems in socioeconomic sphere, and it provides a common framework for resource allocation and decision making of rational individual agents. With the development of Information technology and Internet, dynamic auction has become a research hotspot and gained more and more attention from scholars. How to allocate the varied user concerns in searching to constantly changing online advertisers effectively is the focal issue of online advertising auction. For example, the online booking system needs to deal with the random buyers and the time-sensitive tickets. The distance education system must have dynamic recovery of education resources and be able to handle random remote users’application. Meanwhile, it tries to achieve the optimal general welfare of the system service. Therefore, dynamic auction is widely used and has become a research hotspot in auction.Traditional research mainly concerns about the static auction theory, which has successfully applied to solve a wide range of problems. However, it is inapplicable in dynamic situation which has special properties. Thus both the classical mechanism design of static auction and dynamic factors need to be considered in dynamic auction research.There are two factors in dynamic auction:population dynamic and information dynamic. The former indicates the population of bidders may vary over time and the latter indicates the private information changes. Most of the recent research at home and abroad focuses on the two factors respectively and further refines the mechanism design in specific context.This paper combines the two factors--dynamic population and dynamic information to design mechanism instead of considering only one factor. Besides, because of the Markov property of dynamic auction, Markov process is applied to formalize the description of it, and the expression of auction elements is defined based on the formalization description. In addition, VCG is the classical incentive-compatible static auction mechanism. As a result, based on the DPDI(Dynamic Population&Dynamic Information) dynamic auction model, we combine the static VCG(Vickrey-Clarke-Groves) mechanism and dynamic property of dynamic auction to design the DPDI dynamic auction mechanism, which is incentive-compatible and social welfare maximized. The incentive compatibility is also verified in this paper. As for the mechanism Implementation, the DPDI dynamic auction algorithm is realized for the differentiated DPDI dynamic auction scene, and KPrice algorithm is improved to adapt the DPDI dynamic scene. Finally, after simulating the two implementation methods with random test data, we analyze the experiment result with the key indicators of dynamic auction.This paper explores the dynamic auction scenario of DPDI, constructs a more general DPDI dynamic auction model, and makes the dynamic auction mechanism research more comprehensive. Furthermore, we design DPDI dynamic auction mechanism according to the two dynamic properties, realize the effective implementation of the dynamic auction, satisfy the needs of both bidders and auctioneer, and maintain the equilibrium of the dynamic auction system; Finally, the simulation of dynamic auction scenario and experimental data analysis are also conducted to provide adjustment direction for the optimization of the auction system.