vix.ing · top · new · best · stats · spec

Designing Truthful Contextual Multi-Armed Bandits based Sponsored Search\n Auctions

2020/02/26 by Abhishek Kumar, Abhishek, Kumar, Shweta Jain +3 · 1 citation
Computer Science · Decision Sciences · #Advanced Bandit Algorithms Research #Auction Theory and Applications #Computer Science and Game Theory (cs.GT) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Optimization and Search Problems

paper · pdf · doi:10.48550/arxiv.2002.11349

openalex publication_date 2020/02/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

For sponsored search auctions, we consider contextual multi-armed bandit\nproblem in the presence of strategic agents. In this setting, at each round, an\nadvertising platform (center) runs an auction to select the best-suited ads\nrelevant to the query posted by the user. It is in the best interest of the\ncenter to select an ad that has a high expected value (i.e., probability of\ngetting a click \× value it derives from a click of the ad). The\nprobability of getting a click (CTR) is unknown to the center and depends on\nthe user's profile (context) posting the query. Further, the value derived for\na click is the private information to the advertiser and thus needs to be\nelicited truthfully. The existing solution in this setting is not practical as\nit suffers from very high regret (O(T\(2)/(3))).\n

Cited by

Related