2021/02/27 by Mete Şeref Ahunbay, Ahunbay, Mete Şeref, Adrian Vetta +1
Business, Management and Accounting · Decision Sciences · Economics, Econometrics and Finance · #Auction Theory and Applications #Computer Science and Game Theory (cs.GT) #Economic theories and models #FOS: Computer and information sciences #Supply Chain and Inventory Management
paper · pdf · doi:10.48550/arxiv.2103.00235
openalex publication_date 2021/02/27 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
We study the performance of the Empirical Revenue Maximizing (ERM) mechanism\nin a single-item, single-seller, single-buyer setting. We assume the buyer's\nvaluation is drawn from a regular distribution F and that the seller has\naccess to em two independently drawn samples from F. By solving a family\nof mixed-integer linear programs (MILPs), the ERM mechanism is proven to\nguarantee at least .5914 times the optimal revenue in expectation. Using\nsolutions to these MILPs, we also show that the worst-case efficiency of the\nERM mechanism is at most .61035 times the optimal revenue. These guarantees\nimprove upon the best known lower and upper bounds of .558 and .642,\nrespectively, of [Daskalakis & Zampetakis, '20].\n