When Can the Defender Effectively Deceive Attackers in Security Games?

Authors: Thanh Nguyen, Haifeng Xu9405-9412

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conduct extensive experiments to illustrate our theoretical results in various game settings. Our empirical results align with our theoretical findings, which show that the defender obtains a significant benefit while the attacker suffers a significant loss due to the defender s deception when the attacker plays the Ignorant or Maximin strategies.
Researcher Affiliation Academia Thanh Nguyen1 and Haifeng Xu2 1Department of Computer and Information Science , University of Oregon, USA 2Department of Computer Science, University of Virginia, USA thanhhng@cs.uoregon.edu, hx4ad@virginia.edu
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper does not provide any statement or link indicating that the source code for the described methodology is open source or publicly available.
Open Datasets No The paper states: 'Our experiments use the standard covariance game generator GAMMUT (http: //gamut.stanford/edu), to generate payoff matrices.' It does not use or provide access information for a publicly available or open dataset.
Dataset Splits No The paper does not specify any training, validation, or test dataset splits. It describes generating game instances for experiments.
Hardware Specification Yes Our experiments are conducted on a High Performance Computing (HPC) cluster, with processors are dual E52690v4 (28 cores) and 128 GB memory.
Software Dependencies No The paper mentions 'Cplex to solve our optimization programs' and 'GAMMUT (http: //gamut.stanford/edu)' but does not provide specific version numbers for these software components.
Experiment Setup Yes We consider two cases of the defender s deception capability: (i) small deception interval, i.e., αi =βi =0.05; and big deception interval, i.e., αi =βi = 0.15. In Figure 1(a-b), the defender s utility increases while the attacker s utility decreases gradually as the ratio (k/n) increases.