arXiv - CS - Computer Science and Game Theory最新文献_第3页

Subsidy design for better social outcomes 为取得更好的社会成果而进行补贴设计

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-09-04 DOI: arxiv-2409.03129

Maria-Florina Balcan, Matteo Pozzi, Dravyansh Sharma

{"title":"Subsidy design for better social outcomes","authors":"Maria-Florina Balcan, Matteo Pozzi, Dravyansh Sharma","doi":"arxiv-2409.03129","DOIUrl":"https://doi.org/arxiv-2409.03129","url":null,"abstract":"Overcoming the impact of selfish behavior of rational players in multiagent\u0000systems is a fundamental problem in game theory. Without any intervention from\u0000a central agent, strategic users take actions in order to maximize their\u0000personal utility, which can lead to extremely inefficient overall system\u0000performance, often indicated by a high Price of Anarchy. Recent work (Lin et\u0000al. 2021) investigated and formalized yet another undesirable behavior of\u0000rational agents, that of avoiding freely available information about the game\u0000for selfish reasons, leading to worse social outcomes. A central planner can\u0000significantly mitigate these issues by injecting a subsidy to reduce certain\u0000costs associated with the system and obtain net gains in the system\u0000performance. Crucially, the planner needs to determine how to allocate this\u0000subsidy effectively. We formally show that designing subsidies that perfectly optimize the social\u0000good, in terms of minimizing the Price of Anarchy or preventing the information\u0000avoidance behavior, is computationally hard under standard complexity theoretic\u0000assumptions. On the positive side, we show that we can learn provably good\u0000values of subsidy in repeated games coming from the same domain. This\u0000data-driven subsidy design approach avoids solving computationally hard\u0000problems for unseen games by learning over polynomially many games. We also\u0000show that optimal subsidy can be learned with no-regret given an online\u0000sequence of games, under mild assumptions on the cost matrix. Our study focuses\u0000on two distinct games: a Bayesian extension of the well-studied fair\u0000cost-sharing game, and a component maintenance game with engineering\u0000applications.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"98 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197514","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}

引用次数: 0

Proportionality for Constrained Public Decisions 受限公共决策的相称性

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-09-04 DOI: arxiv-2409.02609

Julian Chingoma, Umberto Grandi, Arianna Novaro

引用次数: 0

Designing Fair Systems for Consumers to Exploit Personalized Pricing 为消费者设计利用个性化定价的公平系统

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-09-04 DOI: arxiv-2409.02777

Aditya Karan, Naina Balepur, Hari Sundaram

{"title":"Designing Fair Systems for Consumers to Exploit Personalized Pricing","authors":"Aditya Karan, Naina Balepur, Hari Sundaram","doi":"arxiv-2409.02777","DOIUrl":"https://doi.org/arxiv-2409.02777","url":null,"abstract":"Many online marketplaces personalize prices based on consumer attributes.\u0000Since these prices are private, consumers will not realize if they spend more\u0000on a good than the lowest possible price, and cannot easily take action to get\u0000better prices. In this paper we introduce a system that takes advantage of\u0000personalized pricing so consumers can profit while improving fairness. Our\u0000system matches consumers for trading; the lower-paying consumer buys the good\u0000for the higher-paying consumer for some fee. We explore various modeling\u0000choices and fairness targets to determine which schema will leave consumers\u0000best off, while also earning revenue for the system itself. We show that when\u0000consumers individually negotiate the transaction price, they are able to\u0000achieve the most fair outcomes. Conversely, when transaction prices are\u0000centrally set, consumers are often unwilling to transact. Minimizing the\u0000average price paid by an individual or group is most profitable for the system,\u0000while achieving a $67%$ reduction in prices. We see that a high dispersion (or\u0000range) of original prices is necessary for our system to be viable. Higher\u0000dispersion can actually lead to increased consumer welfare, and act as a check\u0000against extreme personalization. Our results provide theoretical evidence that\u0000such a system could improve fairness for consumers while sustaining itself\u0000financially.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"17 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197507","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}

引用次数: 0

Achieving Maximin Share and EFX/EF1 Guarantees Simultaneously 同时实现最大份额和 EFX/EF1 保证

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-09-03 DOI: arxiv-2409.01963

Hannaneh Akrami, Nidhi Rathi

{"title":"Achieving Maximin Share and EFX/EF1 Guarantees Simultaneously","authors":"Hannaneh Akrami, Nidhi Rathi","doi":"arxiv-2409.01963","DOIUrl":"https://doi.org/arxiv-2409.01963","url":null,"abstract":"We study the problem of computing emph{fair} divisions of a set of\u0000indivisible goods among agents with emph{additive} valuations. For the past\u0000many decades, the literature has explored various notions of fairness, that can\u0000be primarily seen as either having emph{envy-based} or emph{share-based}\u0000lens. For the discrete setting of resource-allocation problems, emph{envy-free\u0000up to any good} (EFX) and emph{maximin share} (MMS) are widely considered as\u0000the flag-bearers of fairness notions in the above two categories, thereby\u0000capturing different aspects of fairness herein. Due to lack of existence\u0000results of these notions and the fact that a good approximation of EFX or MMS\u0000does not imply particularly strong guarantees of the other, it becomes\u0000important to understand the compatibility of EFX and MMS allocations with one\u0000another. In this work, we identify a novel way to simultaneously achieve MMS\u0000guarantees with EFX/EF1 notions of fairness, while beating the best known\u0000approximation factors [Chaudhury et al., 2021, Amanatidis et al., 2020]. Our\u0000main contribution is to constructively prove the existence of (i) a partial\u0000allocation that is both $2/3$-MMS and EFX, and (ii) a complete allocation that\u0000is both $2/3$-MMS and EF1. Our algorithms run in pseudo-polynomial time if the\u0000approximation factor for MMS is relaxed to $2/3-varepsilon$ for any constant\u0000$varepsilon > 0$ and in polynomial time if, in addition, the EFX (or EF1)\u0000guarantee is relaxed to $(1-delta)$-EFX (or $(1-delta)$-EF1) for any constant\u0000$delta>0$. In particular, we improve from the best approximation factor known\u0000prior to our work, which computes partial allocations that are $1/2$-MMS and\u0000EFX in pseudo-polynomial time [Chaudhury et al., 2021].","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"30 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197511","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}

引用次数: 0

Prophet Inequality from Samples: Is the More the Merrier? 从样本看先知不平等：越多越好吗？

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-08-31 DOI: arxiv-2409.00559

Tomer Ezra

引用次数: 0

Skill Dominance Analysis of Two(Four) player, Four(Five) dice Variant of the Ludo Game 二（四）人四（五）骰子鲁多游戏变体的技能优势分析

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-08-31 DOI: arxiv-2409.00376

Tathagata Banerjee, Diganta Mukherjee

{"title":"Skill Dominance Analysis of Two(Four) player, Four(Five) dice Variant of the Ludo Game","authors":"Tathagata Banerjee, Diganta Mukherjee","doi":"arxiv-2409.00376","DOIUrl":"https://doi.org/arxiv-2409.00376","url":null,"abstract":"This paper examines two different variants of the Ludo game, involving\u0000multiple dice and a fixed number of total turns. Within each variant, multiple\u0000game lengths (total no. of turns) are considered. To compare the two variants,\u0000a set of intuitive, rule-based strategies is designed, representing different\u0000broad methods of strategic play. Game play is simulated between bots (automated\u0000software applications executing repetitive tasks over a network) following\u0000these strategies. The expected results are computed using certain game\u0000theoretic and probabilistic explanations, helping to understand the performance\u0000of the different strategies. The different strategies are further analyzed\u0000using win percentage in a large number of simulations, and Nash Equilibrium\u0000strategies are computed for both variants for a varying number of total turns.\u0000The Nash Equilibrium strategies across different game lengths are compared. A\u0000clear distinction between performances of strategies is observed, with more\u0000sophisticated strategies beating the naive one. A gradual shift in optimal\u0000strategy profiles is observed with changing game length, and certain\u0000sophisticated strategies even confound each other's performance while playing\u0000against each other.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197513","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}

引用次数: 0

PRADA: Proactive Risk Assessment and Mitigation of Misinformed Demand Attacks on Navigational Route Recommendations PRADA：主动风险评估和减少对导航路线建议的误导需求攻击

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-08-30 DOI: arxiv-2409.00243

Ya-Ting Yang, Haozhe Lei, Quanyan Zhu

{"title":"PRADA: Proactive Risk Assessment and Mitigation of Misinformed Demand Attacks on Navigational Route Recommendations","authors":"Ya-Ting Yang, Haozhe Lei, Quanyan Zhu","doi":"arxiv-2409.00243","DOIUrl":"https://doi.org/arxiv-2409.00243","url":null,"abstract":"Leveraging recent advances in wireless communication, IoT, and AI,\u0000intelligent transportation systems (ITS) played an important role in reducing\u0000traffic congestion and enhancing user experience. Within ITS, navigational\u0000recommendation systems (NRS) are essential for helping users simplify route\u0000choices in urban environments. However, NRS are vulnerable to information-based\u0000attacks that can manipulate both the NRS and users to achieve the objectives of\u0000the malicious entities. This study aims to assess the risks of misinformed\u0000demand attacks, where attackers use techniques like Sybil-based attacks to\u0000manipulate the demands of certain origins and destinations considered by the\u0000NRS. We propose a game-theoretic framework for proactive risk assessment of\u0000demand attacks (PRADA) and treat the interaction between attackers and the NRS\u0000as a Stackelberg game. The attacker is the leader who conveys misinformed\u0000demands, while the NRS is the follower responding to the provided information.\u0000Specifically, we consider the case of local-targeted attacks, in which the\u0000attacker aims to make the NRS recommend the authentic users towards a specific\u0000road that favors certain groups. Our analysis unveils the equivalence between\u0000users' incentive compatibility and Wardrop equilibrium recommendations and\u0000shows that the NRS and its users are at high risk when encountering intelligent\u0000attackers who can significantly alter user routes by strategically fabricating\u0000non-existent demands. To mitigate these risks, we introduce a trust mechanism\u0000that leverages users' confidence in the integrity of the NRS, and show that it\u0000can effectively reduce the impact of misinformed demand attacks. Numerical\u0000experiments are used to corroborate the results and demonstrate a Resilience\u0000Paradox, where locally targeted attacks can sometimes benefit the overall\u0000traffic conditions.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"12 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197515","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}

引用次数: 0

Adaptive Incentive-Compatible Navigational Route Recommendations in Urban Transportation Networks 城市交通网络中的自适应激励兼容导航路线建议

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-08-30 DOI: arxiv-2409.00236

Ya-Ting Yang, Haozhe Lei, Quanyan Zhu

{"title":"Adaptive Incentive-Compatible Navigational Route Recommendations in Urban Transportation Networks","authors":"Ya-Ting Yang, Haozhe Lei, Quanyan Zhu","doi":"arxiv-2409.00236","DOIUrl":"https://doi.org/arxiv-2409.00236","url":null,"abstract":"In urban transportation environments, drivers often encounter various path\u0000(route) options when navigating to their destinations. This emphasizes the\u0000importance of navigational recommendation systems (NRS), which simplify\u0000decision-making and reduce travel time for users while alleviating potential\u0000congestion for broader societal benefits. However, recommending the shortest\u0000path may cause the flash crowd effect, and system-optimal routes may not always\u0000align the preferences of human users, leading to non-compliance issues. It is\u0000also worth noting that universal NRS adoption is impractical. Therefore, in\u0000this study, we aim to address these challenges by proposing an incentive\u0000compatibility recommendation system from a game-theoretic perspective and\u0000accounts for non-user drivers with their own path choice behaviors.\u0000Additionally, recognizing the dynamic nature of traffic conditions and the\u0000unpredictability of accidents, this work introduces a dynamic NRS with parallel\u0000and random update schemes, enabling users to safely adapt to changing traffic\u0000conditions while ensuring optimal total travel time costs. The numerical\u0000studies indicate that the proposed parallel update scheme exhibits greater\u0000effectiveness in terms of user compliance, travel time reduction, and\u0000adaptability to the environment.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"43 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142197516","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}

引用次数: 0

Non-uniformly Stable Matchings 非均匀稳定匹配

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-08-29 DOI: arxiv-2408.16271

Naoyuki Kamiyama

引用次数: 0

On the impact of coordinated fleets size on traffic efficiency 协调车队规模对交通效率的影响

arXiv - CS - Computer Science and Game Theory Pub Date : 2024-08-28 DOI: arxiv-2408.15742

Tommaso Toso, Francesca Parise, Paolo Frasca, Alain Y. Kibangou

引用次数: 0