Adversarial risk analysis for bi-agent influence diagrams: An algorithmic approach

González‐Ortega, Jorge; Insua, David Rı́os; Cano, Javier

doi:10.1016/j.ejor.2018.09.015

Cited by 19 publications

(8 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, we have focused on the sequential two stage game problem. The ideas could be extended to other types of games, like the simultaneous Defend-Attack one (Rios and Rios Insua 2012) or the general BAIDs in González-Ortega et al (2019).…”

Section: Discussionmentioning

confidence: 99%

Augmented Probability Simulation Methods for Non-cooperative Games

Ekin,

Naveiro,

Torres-Barrán

et al. 2019

Preprint

View full text Add to dashboard Cite

We present a comprehensive robust decision support framework with novel computational algorithms for decision makers in a non-cooperative sequential setup. Existing simulation based approaches such as Monte Carlo methods can be inefficient under certain conditions like in presence of a high number of decision alternatives and uncertain outcomes. Hence, we provide a novel augmented probability simulation alternative to solve non-cooperative sequential games. We cover approaches to approximate subgame perfect equilibria under common knowledge conditions, assess the robustness of such solutions and, finally, approximate adversarial risk analysis solutions when lacking common knowledge. The proposed approach can be especially beneficial in application domains such as cybersecurity and counter-terrorism.

show abstract

Section: Discussionmentioning

confidence: 99%

Augmented Probability Simulation Methods for Non-cooperative Games

Ekin,

Naveiro,

Torres-Barrán

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…Similarly, nodes D 1 -A 2 reproduce the backbone structure of a Sim D-A template. In González-Ortega et al (2019), a scheme for dealing with these structures was introduced adapting strategic relevance concepts from computational game theory (Koller, 2003), ARA methods (Banks et al, 2015), and classic influence diagram reductions from Shachter (1986). However, the basic reduction operations, were essentially analytic.…”

Section: ¯+mentioning

confidence: 99%

“…The methods in González-Ortega et al (2019) assess analytically a BAID from an ARA perspective. First, they are used to decide whether we deal with a sequential or a simultaneous block.…”

Section: Simulation Based Baid Reduction Operationsmentioning

confidence: 99%

Augmented probability simulation for adversarial risk analysis in general security games

Naveiro¹,

Insua²,

Camacho³

2022

Proceedings of the International Defense and Homeland Security Simulation Workshop

View full text Add to dashboard Cite

Although widely used, standard game-theoretic approaches to security games face severe shortcomings, being the common knowledge assumption a critical one. Adversarial Risk Analysis (ARA) is an alternative modeling framework that mitigates such limitations. However, from a computational perspective, ARA is much more involved than its game theoretical counterparts. We propose an approach for finding ARA solutions to security games represented as bi-agent influence diagrams that is based on augmented probability simulation. We motivate this approach using two simple cases: sequential and simultaneous defend-attack models. We next provide the general framework and illustrate it in handling risks in a cybersecurity setting.

show abstract

“…One can improve this general approach by taking advantage of the dynamic and informational structure of the BAID underlying the problem. One alternates between simulation and optimization as described in González-Ortega et al (2019). But one still needs to simulate and optimize which can be computationally intensive.…”

Section: An Ara Computational Pipelinementioning

confidence: 99%

Adversarial Risk Analysis (Overview)

Banks¹,

Gallego²,

Naveiro³

et al. 2020

Preprint

View full text Add to dashboard Cite

Adversarial risk analysis (ARA) is a relatively new area of research that informs decision-making when facing intelligent opponents and uncertain outcomes. It enables an analyst to express her Bayesian beliefs about an opponent's utilities, capabilities, probabilities and the type of strategic calculation that the opponent is using.Within that framework, the analyst then solves the problem from the perspective of the opponent while placing subjective probability distributions on all unknown quantities. This produces a distribution over the actions of the opponent that permits the analyst to maximize her expected utility. This overview covers conceptual, modeling, computational and applied issues in ARA.

show abstract

Adversarial risk analysis for bi-agent influence diagrams: An algorithmic approach

Cited by 19 publications

References 31 publications

Augmented Probability Simulation Methods for Non-cooperative Games

Augmented Probability Simulation Methods for Non-cooperative Games

Augmented probability simulation for adversarial risk analysis in general security games

Adversarial Risk Analysis (Overview)

Contact Info

Product

Resources

About