Deng, Xiaotie; Zhu, Keyu On Bayesian epistemology of Myerson auction. (English) Zbl 1446.68137 Chen, Jianer (ed.) et al., Frontiers in algorithmics. 12th international workshop, FAW 2018, Guangzhou, China, May 8–10, 2018. Proceedings. Cham: Springer. Lect. Notes Comput. Sci. 10823, 183-196 (2018). Summary: Bayesian Epistemology bases its analysis of the objects under study on a prior, a probability distribution, which is in turn the subject matter in statistical learning, and that of machine learning at least implicitly. We are interested in a game setting where the agents to be learned may shift in accordance with the data collector’s strategies. We focus on this issue of learning and exploiting for Myerson auction where a seller wants to gain information on bidders’ value distributions to achieve the maximum revenue. We show that a world of the power-law distribution would enable the auctioneer to achieve both but the bidders can consistently lie about their probability distribution to improve utility under the other distributions.For the entire collection see [Zbl 1408.68012]. MSC: 68T05 Learning and adaptive systems in artificial intelligence 68T37 Reasoning under uncertainty in the context of artificial intelligence 91A10 Noncooperative games 91A80 Applications of game theory 91B26 Auctions, bargaining, bidding and selling, and other market models Keywords:statistical learning; Bayesian epistemology; probability distribution cheating; Myerson auction; bidding game; Nash equilibrium PDFBibTeX XMLCite \textit{X. Deng} and \textit{K. Zhu}, Lect. Notes Comput. Sci. 10823, 183--196 (2018; Zbl 1446.68137) Full Text: DOI