A computational framework to solve general security games through a Bayesian approach

J. M. Camacho Rodriguez, R. Naveiro, D. Rios Insua

Game-theoretic methods encounter several limitations when dealing with general security games, with common knowledge being a very significant one. Adversarial Risk Analysis (ARA) offers an alternative approach based on Bayesian Decision Analysis that addresses this shortcoming, although it is more computationally involved. To handle general security games computationally, we present an augmented probability simulation scheme to find ARA solutions. We utilize two simple game templates, namely sequential and simultaneous defend-attack models, to demonstrate the general approach, which employs bi-agent influence diagrams as the underlying problem structure. Subsequently, we present a case study to illustrate the proposed approach.

Keywords: Security games, Bayesian Decision Analysis, Augmented Probability Simulation

GT11.BAYES3 Invited Session
November 9, 2023  4:50 PM
HC4: Sacristía Room

M. Kalli, J. Griffin

M. R. Sillero Denamiel, P. Ramirez Cobo, B. Vidakovic

M. Martínez Pizarro, M. Parra Arévalo, E. López Sanjuán, J. Martín Jiménez