Stackelberg games for adversarial prediction problems

By Michael Brückner, Tobias Scheffer. 2011

289 citations recorded by the source corpus.

Subfields: Safety, Robustness & Interpretability, Game & Decision Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Safety, Robustness & Interpretability

In Game & Decision Theory

Citations in the corpus

Adversarial Example Games (2020, neutral): “Such an application falls in line with other general approaches to red teaming AI systems [10] and verifiability in AI development.”
Towards the Science of Security and Privacy in Machine Learning (2016, background): “[92] add a regularizer for the learner and the cost function c(x, x′) as the l2 distance between x and x′.”
Certified Defenses for Data Poisoning Attacks (2017, methodology): “Bilevel optimization has been considered in the related setting of Stackelberg games (Brückner and Scheffer, 2011; Brückner et al., 2012; Zhou and Kantarcioglu, 2016), and is natural to apply here as well.”
Evasion Attacks against Machine Learning at Test Time (2013, methodology): “Under certain conditions, such problems can be solved using a Nash equilibrium approach [7, 8].”
When Does Machine Learning FAIL? Generalized Transferability for Evasion and Poisoning Attacks (2018, background): “Furthermore, [27, 7] introduce game theoretical Stackelberg formulations for the interaction between the adversary and the data miner in the case of data manipulations.”
On the (Statistical) Detection of Adversarial Examples (2017, methodology): “Other approaches make use of game theory [6, 9, 21].”
Data Poisoning Attacks against Autoregressive Models (2016, background): “A somewhat orthogonal line of research has posed the problem of learning in the presence of adversaries in a game theoretic context (Liu and Chawla 2009; Br¨uckner, Kanzow, and Scheffer 2012; Dalvi et al. 2004; Br¨uckner and Scheffer 2009; 2011).”
SoK: Security and Privacy in Machine Learning (2018, background): “iﬁer) and the adversary responds optimally (choose a least cost evasion). Stackelberg games also allow for scalability compared to the corresponding simultaneous move game [90], [91]. Bruckner et al. [92] ﬁrst recognized that test data manipulation can be seen in the PAC framework as modifying the d”
Explicit Defense Actions Against Test-Set Attacks (2017, background): “A separate line of research has posed the problem of learning in the presence of adversaries in game theoretic contexts (Liu and Chawla 2009; Brückner, Kanzow, and Scheffer 2012; Dalvi et al. 2004; Brückner and Scheffer 2009; 2011; Hardt et al. 2016; Brückner and Scheffer 2011).”
Scalable Optimization of Randomized Operational Decisions in Adversarial Classification Settings (2015, methodology): “This decision (and the choice of the threshold) are often motivated by overall tolerance for false positives, as well as operational considerations, for example, to ensure that the number of alerts does not exceed what can reasonably be inspected by security professionals.”
Pattern Recognition Systems under Attack: Design Issues and Research Challenges (2014, background): “[31]; then, more rigorous approaches have introduced Nash and Stackelberg games for secure learning, deriving formal conditions for the corresponding equilibrium strategies [25, 45, 49, 26, 24].”
A Survey on Security Threats and Defensive Techniques of Machine Learning: A Data Driven View (2018, methodology): “[86] M. Brückner, C. Kanzow, and T. Scheffer, ‘‘Static prediction games for adversarial learning problems,’’ J. Mach.”
Hardening Classifiers against Evasion: the Good, the Bad, and the Ugly (2017, background): “Instead, these consider the interaction as a non-zero-sum game, either played simultaneously between the learner and the attacker, or a Stackelberg game, in which the learner is the leader, while the attacker the follower [4], [18], [16], [27].”
Optimal randomized classification in adversarial settings (2014, background): “The study of machine learning in such settings has come to be known as adversarial machine learning [8, 16, 3, 5, 15, 2, 6, 14, 10, 7, 18].”
Optimal Decision Making Under Strategic Behavior (2019, background): “With this motivation, there has been a recent ﬂurry of work on strategic classiﬁcation [1, 2, 5, 6, 8, 11, 16].”
Curie: A method for protecting SVM Classifier from Poisoning Attack (2016, background): “Most of the works [8,7] on secure classiﬁer algorithms uses a game theory model - the problem is modeled as a game between the adversary and the learner.”
Sparse Feature Attacks in Adversarial Learning (2014, methodology): “[10] also made the unstated assumption that the adversary has the ability to change all the features i.”
Convex Adversarial Collective Classification (2013, background): “Dalvi et al. (2004) note that finding a Nash equilibrium is often intractable and propose a strategy to anticipate the adversary’s next move instead.”