We propose a new variant of the strategic classification problem: a principal reveals a classifier, and $n$ agents report their (possibly manipulated) features to be classified. Motivated by real-world applications, our model crucially allows the manipulation of one agent to affect another; that is, it explicitly captures inter-agent externalities. The principal-agent interactions are formally modeled as a Stackelberg game, with the resulting agent manipulation dynamics captured as a simultaneous game. We show that under certain assumptions, the pure Nash Equilibrium of this agent manipulation game is unique and can be efficiently computed. Leveraging this result, PAC learning guarantees are established for the learner: informally, we show that it is possible to learn classifiers that minimize loss on the distribution, even when a random number of agents are manipulating their way to a pure Nash Equilibrium. We also comment on the optimization of such classifiers through gradient-based approaches. This work sets the theoretical foundations for a more realistic analysis of classifiers that are robust against multiple strategic actors interacting in a common environment.
Princeton University（普林斯顿大学）；University of Oregon（俄勒冈大学）；University of Chicago（芝加哥大学）
Most algorithmic studies on multi-agent information design so far have focused on the restricted situation with no inter-agent externalities; a few exceptions investigated truly strategic games such as zero-sum games and second-price auctions but have all focused only on optimal public signaling. This paper initiates the algorithmic information design of both \emph{public} and \emph{private} signaling in a fundamental class of games with negative externalities, i.e., singleton congestion games, with wide application in today's digital economy, machine scheduling, routing, etc. For both public …
University of Waterloo（滑铁卢大学）；York University（约克大学）
Strategic classification, i.e. classification under possible strategic manipulations of features, has received a lot of attention from both the machine learning and the game theory community. Most works focus on analysing properties of the optimal decision rule under such manipulations. In our work we take a learning theoretic perspective, focusing on the sample complexity needed to learn a good decision rule which is robust to strategic manipulation. We perform this analysis by introducing a novel loss function, the \emph{strategic manipulation loss}, which takes into account both the accurac…
Khoury College of Computer Science, Northeastern University（东北大学库利计算机学院）；University of Virginia（弗吉尼亚大学）；Biocomplexity Institute and Initiative, University of Virginia（弗吉尼亚大学生物复杂性研究所）
The study of strategic or adversarial manipulation of testing data to fool a classifier has attracted much recent attention. Most previous works have focused on two extreme situations where any testing data point either is completely adversarial or always equally prefers the positive label. In this paper, we generalize both of these through a unified framework for strategic classification, and introduce the notion of strategic VC-dimension (SVC) to capture the PAC-learnability in our general strategic setup. SVC provably generalizes the recent concept of adversarial VC-dimension (AVC) introduc…
Informatics Institute, University of Amsterdam, The Netherlands（阿姆斯特丹大学信息学院）；ING Research Project Lead（ING研究项目负责人）；ING Analytics, ING Group, Amsterdam, The Netherlands（ING分析部、ING集团、荷兰阿姆斯特丹）；Delft Institute of Applied Mathematics (DIAM), TU Delft, The Netherlands（代尔夫特应用数学研究所（DIAM）、代尔夫特理工大学、荷兰）
We study a class of two-player repeated games with incomplete information and informational externalities. In these games, two states are chosen at the outset, and players get private information on the pair, before engaging in repeated play. The payoff of each player only depends on his `own' state and on his own action. We study to what extent, and how, information can be exchanged in equilibrium. We prove that provided the private information of each player is valuable for the other player, the set of sequential equilibrium payoffs converges to the set of feasible and individually rational …
IBM India Research Lab（IBM印度研究院）；Yahoo! Labs（雅虎实验室）；Dept. of CSA, IISc（印度科学理工学院计算机科学与自动化系）
We consider the problem of Probably Approximate Correct (PAC) learning of a binary classifier from noisy labeled examples acquired from multiple annotators (each characterized by a respective classification noise rate). First, we consider the complete information scenario, where the learner knows the noise rates of all the annotators. For this scenario, we derive sample complexity bound for the Minimum Disagreement Algorithm (MDA) on the number of labeled examples to be obtained from each annotator. Next, we consider the incomplete information scenario, where each annotator is strategic and ho…
Persuasion studies how a principal can influence agents' decisions via strategic information revelation --- often described as a signaling scheme --- in order to yield the most desirable equilibrium outcome. Recently, there has been a large body of algorithmic study of designing optimal public signaling schemes, a.k.a., public persuasion, which however is rifle with computational intractability results. In this paper, we design efficient and tight algorithms for public persuasion, and focus on a fundamental multi-agent persuasion model with no inter-agent externalities and binary actions intro…
University of California, Berkeley（加州大学伯克利分校）；University of Texas, Austin（德克萨斯大学奥斯汀分校）；Cornell Tech（康奈尔科技）；Stanford University（斯坦福大学）；University of Pennsylvania（宾夕法尼亚大学）；Harvard University（哈佛大学）；Inria, Paris（巴黎Inria）
We study games in which a network mediates strategic spillovers and externalities among the players. How does a planner optimally target interventions that change individuals' private returns to investment? We analyze this question by decomposing any intervention into orthogonal principal components, which are determined by the network and are ordered according to their associated eigenvalues. There is a close connection between the nature of spillovers and the representation of various principal components in the optimal intervention. In games of strategic complements (substitutes), intervent…
Harvard University（哈佛大学）；UC Santa Cruz（加州大学圣克鲁兹分校）
We address the question of repeatedly learning linear classifiers against agents who are strategically trying to game the deployed classifiers, and we use the Stackelberg regret to measure the performance of our algorithms. First, we show that Stackelberg and external regret for the problem of strategic classification are strongly incompatible: i.e., there exist worst-case scenarios, where any sequence of actions providing sublinear external regret might result in linear Stackelberg regret and vice versa. Second, we present a strategy-aware algorithm for minimizing the Stackelberg regret for w…
Technion - Israel Institute of Technology（以色列理工学院）
Strategic classification regards the problem of learning in settings where users can strategically modify their features to improve outcomes. This setting applies broadly and has received much recent attention. But despite its practical significance, work in this space has so far been predominantly theoretical. In this paper we present a learning framework for strategic classification that is practical. Our approach directly minimizes the "strategic" empirical risk, achieved by differentiating through the strategic response of users. This provides flexibility that allows us to extend beyond th…
The COVID-19 lockdowns have created a significant socioeconomic impact on our society. In this paper, we propose a population vaccination game framework, called EPROACH, to design policies for reopenings that guarantee post-opening public health safety. In our framework, a population of players decides whether to vaccinate or not based on the public and private information they receive. The reopening is captured by the switching of the game state. The insights obtained from our framework include the appropriate vaccination coverage threshold for safe-reopening and information-based methods to …
A digital security breach, by which confidential information is leaked, does not only affect the agent whose system is infiltrated, but is also detrimental to other agents socially connected to the infiltrated system. Although it has been argued that these externalities create incentives to under-invest in security, this presumption is challenged by the possibility of strategic adversaries that attack the least protected agents. In this paper we study a new model of security games in which agents share tokens of sensitive information in a network of contacts. The agents have the opportunity to…
As predictive models are deployed into the real world, they must increasingly contend with strategic behavior. A growing body of work on strategic classification treats this problem as a Stackelberg game: the decision-maker "leads" in the game by deploying a model, and the strategic agents "follow" by playing their best response to the deployed model. Importantly, in this framing, the burden of learning is placed solely on the decision-maker, while the agents' best responses are implicitly treated as instantaneous. In this work, we argue that the order of play in strategic classification is fu…
Non-Linear Strategic Classification Made Practical
School of Informatics, University of Edinburgh（爱丁堡大学信息学院）；Department of Engineering Science, University of Oxford（牛津大学工程科学系）；School of Informatics University of Edinburgh（爱丁堡大学信息学院）
Migrating Birds Optimization-Based Feature Selection for Text Classification
Scientific and Technical Research Council of Turkiye (TUBITAK)（土耳其科学技术研究委员会）；Dogus University（多乌什大学）；Kocaeli University（科贾埃利大学）；Acibadem University（阿奇巴德姆大学）
We study an online linear classification problem, in which the data is generated by strategic agents who manipulate their features in an effort to change the classification outcome. In rounds, the learner deploys a classifier, and an adversarially chosen agent arrives, possibly manipulating her features to optimally respond to the learner. The learner has no knowledge of the agents' utility functions or "real" features, which may vary widely across agents. Instead, the learner is only able to observe their "revealed preferences" --- i.e. the actual manipulated feature vectors they provide. For…
Breaking the Gradient Barrier: Unveiling Large Language Models for Strategic Classification
College of Computer Science and Technology, National University of Defense Technology（国防科技大学计算机科学与技术学院）；Center for Data Science, Peking University（北京大学数据科学中心）；Faculty of Engineering, the University of Sydney（悉尼大学工程学院）；Faculty of Computing, Harbin Institute of Technology（哈尔滨工业大学计算机学院）
Duke University（杜克大学）；University of Southern California（南加州大学）
Machine learning techniques can be useful in applications such as credit approval and college admission. However, to be classified more favorably in such contexts, an agent may decide to strategically withhold some of her features, such as bad test scores. This is a missing data problem with a twist: which data is missing {\em depends on the chosen classifier}, because the specific classifier is what may create the incentive to withhold certain feature values. We address the problem of training classifiers that are robust to this behavior. We design three classification methods: {\sc Mincut}, …
Washington University in Saint Louis（圣路易斯华盛顿大学）；George Mason University（乔治梅森大学）；University of California Santa Cruz（加州大学圣克鲁兹分校）
The use of algorithmic decision making systems in domains which impact the financial, social, and political well-being of people has created a demand for these decision making systems to be "fair" under some accepted notion of equity. This demand has in turn inspired a large body of work focused on the development of fair learning algorithms which are then used in lieu of their conventional counterparts. Most analysis of such fair algorithms proceeds from the assumption that the people affected by the algorithmic decisions are represented as immutable feature vectors. However, strategic agents…
Malicious softwares or malwares for short have become a major security threat. While originating in criminal behavior, their impact are also influenced by the decisions of legitimate end users. Getting agents in the Internet, and in networks in general, to invest in and deploy security features and protocols is a challenge, in particular because of economic reasons arising from the presence of network externalities. In this paper, we focus on the question of incentive alignment for agents of a large network towards a better security. We start with an economic model for a single agent, that det…
When reasoning about strategic behavior in a machine learning context it is tempting to combine standard microfoundations of rational agents with the statistical decision theory underlying classification. In this work, we argue that a direct combination of these standard ingredients leads to brittle solution concepts of limited descriptive and prescriptive value. First, we show that rational agents with perfect information produce discontinuities in the aggregate response to a decision rule that we often do not observe empirically. Second, when any positive fraction of agents is not perfectly …
Toyota Technological Institute at Chicago（芝加哥丰田技术研究所）；Carnegie Mellon University（卡内基梅隆大学）；Northwestern University（西北大学）
In this work, we consider classification of agents who can both game and improve. For example, people wishing to get a loan may be able to take some actions that increase their perceived credit-worthiness and others that also increase their true credit-worthiness. A decision-maker would like to define a classification rule with few false-positives (does not give out many bad loans) while yielding many true positives (giving out many good loans), which includes encouraging agents to improve to become true positives if possible. We consider two models for this problem, a general discrete model a…
In many societal resource allocation domains, machine learning methods are increasingly used to either score or rank agents in order to decide which ones should receive either resources (e.g., homeless services) or scrutiny (e.g., child welfare investigations) from social services agencies. An agency's scoring function typically operates on a feature vector that contains a combination of self-reported features and information available to the agency about individuals or households.This can create incentives for agents to misrepresent their self-reported features in order to receive resources o…
Toyota Technological Institute at Chicago（芝加哥丰田技术研究所）；Georgia Institute of Technology（佐治亚理工学院）
We initiate the study of strategic behavior in screening processes with multiple classifiers. We focus on two contrasting settings: a conjunctive setting in which an individual must satisfy all classifiers simultaneously, and a sequential setting in which an individual to succeed must satisfy classifiers one at a time. In other words, we introduce the combination of strategic classification with screening processes. We show that sequential screening pipelines exhibit new and surprising behavior where individuals can exploit the sequential ordering of the tests to zig-zag between classifiers wi…
When consequential decisions are informed by algorithmic input, individuals may feel compelled to alter their behavior in order to gain a system's approval. Models of agent responsiveness, termed "strategic manipulation," analyze the interaction between a learner and agents in a world where all agents are equally able to manipulate their features in an attempt to "trick" a published classifier. In cases of real world classification, however, an agent's ability to adapt to an algorithm is not simply a function of her personal interest in receiving a positive classification, but is bound up in a…
Technion Israel Institute of Technology（以色列理工学院）；Indian Institute of Technology Hyderabad（印度理工学院海得拉巴分校）；Google Research（谷歌研究院）
Humans have come to rely on machines for reducing excessive information to manageable representations. But this reliance can be abused -- strategic machines might craft representations that manipulate their users. How can a user make good choices based on strategic representations? We formalize this as a learning problem, and pursue algorithms for decision-making that are robust to manipulation. In our main setting of interest, the system represents attributes of an item to the user, who then decides whether or not to consume. We model this interaction through the lens of strategic classificat…
We consider a game in which a strategic defender classifies an intruder as spy or spammer. The classification is based on the number of file server and mail server attacks observed during a fixed window. The spammer naively attacks (with a known distribution) his main target: the mail server. The spy strategically selects the number of attacks on his main target: the file server. The defender strategically selects his classification policy: a threshold on the number of file server attacks. We model the interaction of the two players (spy and defender) as a nonzero-sum game: The defender needs …
We study the strategic simplicity of stable matching mechanisms where one side has fixed preferences, termed priorities. Specifically, we ask which priorities are such that the strategyproofness of deferred acceptance (DA) can be recognized by agents unable to perform contingency reasoning, that is, \emph{when is DA obviously strategyproof} (Li, 2017)? We answer this question by completely characterizing those priorities which make DA obviously strategyproof (OSP). This solves an open problem of Ashlagi and Gonczarowski, 2018. We find that when DA is OSP, priorities are either acyclic (Ergin, …
A bidimensional representation of the space of 2x2 Symmetric Games in the strategic representation is proposed. This representation provides a tool for the classification of 2x2 symmetric games, quantification of the fraction of them having a certain feature, and predictions of changes in the characteristics of a game when a change in done on the payoff matrix that defines it.
