Haris Aziz, Xin Huang, Kei Kimura, Indrajit Saha, Zhaohong Sun Mashbat Suzuki, Makoto Yokoo
We consider the problem of fair allocation with subsidy when agents have weighted entitlements. After highlighting several important differences from the unweighted cases, we present several results concerning weighted envy-freeability including general characterizations, algorithms for achieving and testing weighted envy-freeability, lower and upper bounds for worst case subsidy for non-wasteful and envy-freeable allocations, and algorithms for achieving weighted envy-freeability along with other properties.
{"title":"Weighted Envy-free Allocation with Subsidy","authors":"Haris Aziz, Xin Huang, Kei Kimura, Indrajit Saha, Zhaohong Sun Mashbat Suzuki, Makoto Yokoo","doi":"arxiv-2408.08711","DOIUrl":"https://doi.org/arxiv-2408.08711","url":null,"abstract":"We consider the problem of fair allocation with subsidy when agents have\u0000weighted entitlements. After highlighting several important differences from\u0000the unweighted cases, we present several results concerning weighted\u0000envy-freeability including general characterizations, algorithms for achieving\u0000and testing weighted envy-freeability, lower and upper bounds for worst case\u0000subsidy for non-wasteful and envy-freeable allocations, and algorithms for\u0000achieving weighted envy-freeability along with other properties.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197579","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Perpetual voting studies fair collective decision-making in settings where many decisions are to be made, and is a natural framework for settings such as parliaments and the running of blockchain Decentralized Autonomous Organizations (DAOs). We focus our attention on the binary case (YES/NO decisions) and textit{individual} guarantees for each of the participating agents. We introduce a novel notion, inspired by the popular maxi-min-share (MMS) for fair allocation. The agent expects to get as many decisions as if they were to optimally partition the decisions among the agents, with an adversary deciding which of the agents decides on what bundle. We show an online algorithm that guarantees the MMS notion for $n=3$ agents, an offline algorithm for $n=4$ agents, and show that no online algorithm can guarantee the $MMS^{adapt}$ for $ngeq 7$ agents. We also show that the Maximum Nash Welfare (MNW) outcome can only guarantee $O(frac{1}{n})$ of the MMS notion in the worst case.
{"title":"Beyond Proportional Individual Guarantees for Binary Perpetual Voting","authors":"Yotam Gafni, Ben Golan","doi":"arxiv-2408.08767","DOIUrl":"https://doi.org/arxiv-2408.08767","url":null,"abstract":"Perpetual voting studies fair collective decision-making in settings where\u0000many decisions are to be made, and is a natural framework for settings such as\u0000parliaments and the running of blockchain Decentralized Autonomous\u0000Organizations (DAOs). We focus our attention on the binary case (YES/NO\u0000decisions) and textit{individual} guarantees for each of the participating\u0000agents. We introduce a novel notion, inspired by the popular maxi-min-share\u0000(MMS) for fair allocation. The agent expects to get as many decisions as if\u0000they were to optimally partition the decisions among the agents, with an\u0000adversary deciding which of the agents decides on what bundle. We show an\u0000online algorithm that guarantees the MMS notion for $n=3$ agents, an offline\u0000algorithm for $n=4$ agents, and show that no online algorithm can guarantee the\u0000$MMS^{adapt}$ for $ngeq 7$ agents. We also show that the Maximum Nash Welfare\u0000(MNW) outcome can only guarantee $O(frac{1}{n})$ of the MMS notion in the\u0000worst case.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"43 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197557","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We study the problem of no-regret learning algorithms for general monotone and smooth games and their last-iterate convergence properties. Specifically, we investigate the problem under bandit feedback and strongly uncoupled dynamics, which allows modular development of the multi-player system that applies to a wide range of real applications. We propose a mirror-descent-based algorithm, which converges in $O(T^{-1/4})$ and is also no-regret. The result is achieved by a dedicated use of two regularizations and the analysis of the fixed point thereof. The convergence rate is further improved to $O(T^{-1/2})$ in the case of strongly monotone games. Motivated by practical tasks where the game evolves over time, the algorithm is extended to time-varying monotone games. We provide the first non-asymptotic result in converging monotone games and give improved results for equilibrium tracking games.
{"title":"Uncoupled and Convergent Learning in Monotone Games under Bandit Feedback","authors":"Jing Dong, Baoxiang Wang, Yaoliang Yu","doi":"arxiv-2408.08395","DOIUrl":"https://doi.org/arxiv-2408.08395","url":null,"abstract":"We study the problem of no-regret learning algorithms for general monotone\u0000and smooth games and their last-iterate convergence properties. Specifically,\u0000we investigate the problem under bandit feedback and strongly uncoupled\u0000dynamics, which allows modular development of the multi-player system that\u0000applies to a wide range of real applications. We propose a mirror-descent-based\u0000algorithm, which converges in $O(T^{-1/4})$ and is also no-regret. The result\u0000is achieved by a dedicated use of two regularizations and the analysis of the\u0000fixed point thereof. The convergence rate is further improved to $O(T^{-1/2})$\u0000in the case of strongly monotone games. Motivated by practical tasks where the\u0000game evolves over time, the algorithm is extended to time-varying monotone\u0000games. We provide the first non-asymptotic result in converging monotone games\u0000and give improved results for equilibrium tracking games.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"11 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197558","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We study variants of the Optimal Refugee Resettlement problem where a set $F$ of refugee families need to be allocated to a set $L$ of possible places of resettlement in a feasible and optimal way. Feasibility issues emerge from the assumption that each family requires certain services (such as accommodation, school seats, or medical assistance), while there is an upper and, possibly, a lower quota on the number of service units provided at a given place. Besides studying the problem of finding a feasible assignment, we also investigate two natural optimization variants. In the first one, we allow families to express preferences over $P$, and we aim for a Pareto-optimal assignment. In a more general setting, families can attribute utilities to each place in $P$, and the task is to find a feasible assignment with maximum total utilities. We study the computational complexity of all three variants in a multivariate fashion using the framework of parameterized complexity. We provide fixed-parameter tractable algorithms for a handful of natural parameterizations, and complement these tractable cases with tight intractability results.
{"title":"Parameterized Algorithms for Optimal Refugee Resettlement","authors":"Jiehua Chen, Ildikó Schlotter, Sofia Simola","doi":"arxiv-2408.08392","DOIUrl":"https://doi.org/arxiv-2408.08392","url":null,"abstract":"We study variants of the Optimal Refugee Resettlement problem where a set $F$\u0000of refugee families need to be allocated to a set $L$ of possible places of\u0000resettlement in a feasible and optimal way. Feasibility issues emerge from the\u0000assumption that each family requires certain services (such as accommodation,\u0000school seats, or medical assistance), while there is an upper and, possibly, a\u0000lower quota on the number of service units provided at a given place. Besides\u0000studying the problem of finding a feasible assignment, we also investigate two\u0000natural optimization variants. In the first one, we allow families to express\u0000preferences over $P$, and we aim for a Pareto-optimal assignment. In a more\u0000general setting, families can attribute utilities to each place in $P$, and the\u0000task is to find a feasible assignment with maximum total utilities. We study\u0000the computational complexity of all three variants in a multivariate fashion\u0000using the framework of parameterized complexity. We provide fixed-parameter\u0000tractable algorithms for a handful of natural parameterizations, and complement\u0000these tractable cases with tight intractability results.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"98 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Nivasini Ananthakrishnan, Nika Haghtalab, Chara Podimata, Kunhe Yang
When learning in strategic environments, a key question is whether agents can overcome uncertainty about their preferences to achieve outcomes they could have achieved absent any uncertainty. Can they do this solely through interactions with each other? We focus this question on the ability of agents to attain the value of their Stackelberg optimal strategy and study the impact of information asymmetry. We study repeated interactions in fully strategic environments where players' actions are decided based on learning algorithms that take into account their observed histories and knowledge of the game. We study the pure Nash equilibria (PNE) of a meta-game where players choose these algorithms as their actions. We demonstrate that if one player has perfect knowledge about the game, then any initial informational gap persists. That is, while there is always a PNE in which the informed agent achieves her Stackelberg value, there is a game where no PNE of the meta-game allows the partially informed player to achieve her Stackelberg value. On the other hand, if both players start with some uncertainty about the game, the quality of information alone does not determine which agent can achieve her Stackelberg value. In this case, the concept of information asymmetry becomes nuanced and depends on the game's structure. Overall, our findings suggest that repeated strategic interactions alone cannot facilitate learning effectively enough to earn an uninformed player her Stackelberg value.
{"title":"Is Knowledge Power? On the (Im)possibility of Learning from Strategic Interaction","authors":"Nivasini Ananthakrishnan, Nika Haghtalab, Chara Podimata, Kunhe Yang","doi":"arxiv-2408.08272","DOIUrl":"https://doi.org/arxiv-2408.08272","url":null,"abstract":"When learning in strategic environments, a key question is whether agents can\u0000overcome uncertainty about their preferences to achieve outcomes they could\u0000have achieved absent any uncertainty. Can they do this solely through\u0000interactions with each other? We focus this question on the ability of agents\u0000to attain the value of their Stackelberg optimal strategy and study the impact\u0000of information asymmetry. We study repeated interactions in fully strategic\u0000environments where players' actions are decided based on learning algorithms\u0000that take into account their observed histories and knowledge of the game. We\u0000study the pure Nash equilibria (PNE) of a meta-game where players choose these\u0000algorithms as their actions. We demonstrate that if one player has perfect\u0000knowledge about the game, then any initial informational gap persists. That is,\u0000while there is always a PNE in which the informed agent achieves her\u0000Stackelberg value, there is a game where no PNE of the meta-game allows the\u0000partially informed player to achieve her Stackelberg value. On the other hand,\u0000if both players start with some uncertainty about the game, the quality of\u0000information alone does not determine which agent can achieve her Stackelberg\u0000value. In this case, the concept of information asymmetry becomes nuanced and\u0000depends on the game's structure. Overall, our findings suggest that repeated\u0000strategic interactions alone cannot facilitate learning effectively enough to\u0000earn an uninformed player her Stackelberg value.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"68 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197576","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Gagan Aggarwal, Ashwinkumar Badanidiyuru, Santiago R. Balseiro, Kshipra Bhawalkar, Yuan Deng, Zhe Feng, Gagan Goel, Christopher Liaw, Haihao Lu, Mohammad Mahdian, Jieming Mao, Aranyak Mehta, Vahab Mirrokni, Renato Paes Leme, Andres Perlroth, Georgios Piliouras, Jon Schneider, Ariel Schvartzman, Balasubramanian Sivan, Kelly Spendlove, Yifeng Teng, Di Wang, Hanrui Zhang, Mingfei Zhao, Wennan Zhu, Song Zuo
In this survey, we summarize recent developments in research fueled by the growing adoption of automated bidding strategies in online advertising. We explore the challenges and opportunities that have arisen as markets embrace this autobidding and cover a range of topics in this area, including bidding algorithms, equilibrium analysis and efficiency of common auction formats, and optimal auction design.
{"title":"Auto-bidding and Auctions in Online Advertising: A Survey","authors":"Gagan Aggarwal, Ashwinkumar Badanidiyuru, Santiago R. Balseiro, Kshipra Bhawalkar, Yuan Deng, Zhe Feng, Gagan Goel, Christopher Liaw, Haihao Lu, Mohammad Mahdian, Jieming Mao, Aranyak Mehta, Vahab Mirrokni, Renato Paes Leme, Andres Perlroth, Georgios Piliouras, Jon Schneider, Ariel Schvartzman, Balasubramanian Sivan, Kelly Spendlove, Yifeng Teng, Di Wang, Hanrui Zhang, Mingfei Zhao, Wennan Zhu, Song Zuo","doi":"arxiv-2408.07685","DOIUrl":"https://doi.org/arxiv-2408.07685","url":null,"abstract":"In this survey, we summarize recent developments in research fueled by the\u0000growing adoption of automated bidding strategies in online advertising. We\u0000explore the challenges and opportunities that have arisen as markets embrace\u0000this autobidding and cover a range of topics in this area, including bidding\u0000algorithms, equilibrium analysis and efficiency of common auction formats, and\u0000optimal auction design.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"164 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197577","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In many settings, there is an organizer who would like to divide a set of agents into $k$ coalitions, and cares about the friendships within each coalition. Specifically, the organizer might want to maximize utilitarian social welfare, maximize egalitarian social welfare, or simply guarantee that every agent will have at least one friend within his coalition. However, in many situations, the organizer is not familiar with the friendship connections, and he needs to obtain them from the agents. In this setting, a manipulative agent may falsely report friendship connections in order to increase his utility. In this paper, we analyze the complexity of finding manipulation in such $k$-coalitional games on graphs. We also introduce a new type of manipulation, socially-aware manipulation, in which the manipulator would like to increase his utility without decreasing the social welfare. We then study the complexity of finding socially-aware manipulation in our setting. Finally, we examine the frequency of socially-aware manipulation and the running time of our algorithms via simulation results.
{"title":"The Complexity of Manipulation of k-Coalitional Games on Graphs","authors":"Hodaya Barr, Yohai Trabelsi, Sarit Kraus, Liam Roditty, Noam Hazon","doi":"arxiv-2408.07368","DOIUrl":"https://doi.org/arxiv-2408.07368","url":null,"abstract":"In many settings, there is an organizer who would like to divide a set of\u0000agents into $k$ coalitions, and cares about the friendships within each\u0000coalition. Specifically, the organizer might want to maximize utilitarian\u0000social welfare, maximize egalitarian social welfare, or simply guarantee that\u0000every agent will have at least one friend within his coalition. However, in\u0000many situations, the organizer is not familiar with the friendship connections,\u0000and he needs to obtain them from the agents. In this setting, a manipulative\u0000agent may falsely report friendship connections in order to increase his\u0000utility. In this paper, we analyze the complexity of finding manipulation in\u0000such $k$-coalitional games on graphs. We also introduce a new type of\u0000manipulation, socially-aware manipulation, in which the manipulator would like\u0000to increase his utility without decreasing the social welfare. We then study\u0000the complexity of finding socially-aware manipulation in our setting. Finally,\u0000we examine the frequency of socially-aware manipulation and the running time of\u0000our algorithms via simulation results.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"2 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197578","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Determining how close a winner of an election is to becoming a loser, or distinguishing between different possible winners of an election, are major problems in computational social choice. We tackle these problems for so-called weighted tournament solutions by generalizing the notion of margin of victory (MoV) for tournament solutions by Brill et. al to weighted tournament solutions. For these, the MoV of a winner (resp. loser) is the total weight that needs to be changed in the tournament to make them a loser (resp. winner). We study three weighted tournament solutions: Borda's rule, the weighted Uncovered Set, and Split Cycle. For all three rules, we determine whether the MoV for winners and non-winners is tractable and give upper and lower bounds on the possible values of the MoV. Further, we axiomatically study and generalize properties from the unweighted tournament setting to weighted tournaments.
{"title":"Margin of Victory for Weighted Tournament Solutions","authors":"Michelle Döring, Jannik Peters","doi":"arxiv-2408.06873","DOIUrl":"https://doi.org/arxiv-2408.06873","url":null,"abstract":"Determining how close a winner of an election is to becoming a loser, or\u0000distinguishing between different possible winners of an election, are major\u0000problems in computational social choice. We tackle these problems for so-called\u0000weighted tournament solutions by generalizing the notion of margin of victory\u0000(MoV) for tournament solutions by Brill et. al to weighted tournament\u0000solutions. For these, the MoV of a winner (resp. loser) is the total weight\u0000that needs to be changed in the tournament to make them a loser (resp. winner).\u0000We study three weighted tournament solutions: Borda's rule, the weighted\u0000Uncovered Set, and Split Cycle. For all three rules, we determine whether the\u0000MoV for winners and non-winners is tractable and give upper and lower bounds on\u0000the possible values of the MoV. Further, we axiomatically study and generalize\u0000properties from the unweighted tournament setting to weighted tournaments.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"121 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197583","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the era of Web3, decentralized technologies have emerged as the cornerstone of a new digital paradigm. Backed by a decentralized blockchain architecture, the Web3 space aims to democratize all aspects of the web. From data-sharing to learning models, outsourcing computation is an established, prevalent practice. Verifiable computation makes this practice trustworthy as clients/users can now efficiently validate the integrity of a computation. As verifiable computation gets considered for applications in the Web3 space, decentralization is crucial for system reliability, ensuring that no single entity can suppress clients. At the same time, however, decentralization needs to be balanced with efficiency: clients want their computations done as quickly as possible. Motivated by these issues, we study the trade-off between decentralization and efficiency when outsourcing computational tasks to strategic, rational solution providers. Specifically, we examine this trade-off when the client employs (1) revelation mechanisms, i.e. auctions, where solution providers bid their desired reward for completing the task by a specific deadline and then the client selects which of them will do the task and how much they will be rewarded, and (2) simple, non-revelation mechanisms, where the client commits to the set of rules she will use to map solutions at specific times to rewards and then solution providers decide whether they want to do the task or not. We completely characterize the power and limitations of revelation and non-revelation mechanisms in our model.
{"title":"V3rified: Revelation vs Non-Revelation Mechanisms for Decentralized Verifiable Computation","authors":"Tiantian Gong, Aniket Kate, Alexandros Psomas, Athina Terzoglou","doi":"arxiv-2408.07177","DOIUrl":"https://doi.org/arxiv-2408.07177","url":null,"abstract":"In the era of Web3, decentralized technologies have emerged as the\u0000cornerstone of a new digital paradigm. Backed by a decentralized blockchain\u0000architecture, the Web3 space aims to democratize all aspects of the web. From\u0000data-sharing to learning models, outsourcing computation is an established,\u0000prevalent practice. Verifiable computation makes this practice trustworthy as\u0000clients/users can now efficiently validate the integrity of a computation. As\u0000verifiable computation gets considered for applications in the Web3 space,\u0000decentralization is crucial for system reliability, ensuring that no single\u0000entity can suppress clients. At the same time, however, decentralization needs\u0000to be balanced with efficiency: clients want their computations done as quickly\u0000as possible. Motivated by these issues, we study the trade-off between decentralization\u0000and efficiency when outsourcing computational tasks to strategic, rational\u0000solution providers. Specifically, we examine this trade-off when the client\u0000employs (1) revelation mechanisms, i.e. auctions, where solution providers bid\u0000their desired reward for completing the task by a specific deadline and then\u0000the client selects which of them will do the task and how much they will be\u0000rewarded, and (2) simple, non-revelation mechanisms, where the client commits\u0000to the set of rules she will use to map solutions at specific times to rewards\u0000and then solution providers decide whether they want to do the task or not. We\u0000completely characterize the power and limitations of revelation and\u0000non-revelation mechanisms in our model.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"29 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We provide the first analysis of clock auctions through the learning-augmented framework. Deferred-acceptance clock auctions are a compelling class of mechanisms satisfying a unique list of highly practical properties, including obvious strategy-proofness, transparency, and unconditional winner privacy, making them particularly well-suited for real-world applications. However, early work that evaluated their performance from a worst-case analysis standpoint concluded that no deterministic clock auction can achieve much better than an $O(log n)$ approximation of the optimal social welfare (where $n$ is the number of bidders participating in the auction), even in seemingly very simple settings. To overcome this overly pessimistic impossibility result, which heavily depends on the assumption that the designer has no information regarding the preferences of the participating bidders, we leverage the learning-augmented framework. This framework assumes that the designer is provided with some advice regarding what the optimal solution may be. This advice may be the product of machine-learning algorithms applied to historical data, so it can provide very useful guidance, but it can also be highly unreliable. Our main results are learning-augmented clock auctions that use this advice to achieve much stronger performance guarantees whenever the advice is accurate (known as consistency), while simultaneously maintaining worst-case guarantees even if this advice is arbitrarily inaccurate (known as robustness). Specifically, for the standard notion of consistency, we provide a clock auction that achieves the best of both worlds: $(1+epsilon)$-consistency for any constant $epsilon > 0$ and $O(log n)$ robustness. We then also consider a much stronger notion of consistency and provide an auction that achieves the optimal trade-off between this notion of consistency and robustness.
{"title":"Clock Auctions Augmented with Unreliable Advice","authors":"Vasilis Gkatzelis, Daniel Schoepflin, Xizhi Tan","doi":"arxiv-2408.06483","DOIUrl":"https://doi.org/arxiv-2408.06483","url":null,"abstract":"We provide the first analysis of clock auctions through the\u0000learning-augmented framework. Deferred-acceptance clock auctions are a\u0000compelling class of mechanisms satisfying a unique list of highly practical\u0000properties, including obvious strategy-proofness, transparency, and\u0000unconditional winner privacy, making them particularly well-suited for\u0000real-world applications. However, early work that evaluated their performance\u0000from a worst-case analysis standpoint concluded that no deterministic clock\u0000auction can achieve much better than an $O(log n)$ approximation of the\u0000optimal social welfare (where $n$ is the number of bidders participating in the\u0000auction), even in seemingly very simple settings. To overcome this overly pessimistic impossibility result, which heavily\u0000depends on the assumption that the designer has no information regarding the\u0000preferences of the participating bidders, we leverage the learning-augmented\u0000framework. This framework assumes that the designer is provided with some\u0000advice regarding what the optimal solution may be. This advice may be the\u0000product of machine-learning algorithms applied to historical data, so it can\u0000provide very useful guidance, but it can also be highly unreliable. Our main results are learning-augmented clock auctions that use this advice\u0000to achieve much stronger performance guarantees whenever the advice is accurate\u0000(known as consistency), while simultaneously maintaining worst-case guarantees\u0000even if this advice is arbitrarily inaccurate (known as robustness).\u0000Specifically, for the standard notion of consistency, we provide a clock\u0000auction that achieves the best of both worlds: $(1+epsilon)$-consistency for\u0000any constant $epsilon > 0$ and $O(log n)$ robustness. We then also consider a\u0000much stronger notion of consistency and provide an auction that achieves the\u0000optimal trade-off between this notion of consistency and robustness.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"11 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197584","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：