Security Games with Information Leakage: Modeling and Computation

Authors: Haifeng Xu, Albert Xing Jiang, Arunesh Sinha, Zinovi Rabinovich, Shaddin Dughmi, Milind Tambe

IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental 5 Experiments Traditional algorithms for computing Strong Stackelberg Equilibrium (SSE) only optimize the coverage probability at each target, without considering their correlations. In this section, we experimentally study how traditional algorithms and our new algorithms perform in presence of probabilistic or adversarial information leakage. ... All algorithms are tested on the following two sets of data: Los Angeles International Airport (LAX) Checkpoint Data ... Simulated Game Payoffs.
Researcher Affiliation Academia Haifeng Xu Albert X. Jiang Arunesh Sinha USC Trinity University USC haifengx@usc.edu xjiang@trinity.edu aruneshs@usc.edu Zinovi Rabinovich Shaddin Dughmi Milind Tambe Independent Researcher USC USC zr@zinovi.net shaddin@usc.edu tambe@usc.edu
Pseudocode Yes Algorithm 1 Defender Oracle Input: matrix A of form (4). Output: a pure strategy s. ... Algorithm 2 Max-Entropy Sampling Input: : α [0, )n, k. Output: : a pure strategy s with |s| = k.
Open Source Code No The paper does not provide any explicit statements about releasing source code or links to a code repository for the described methodology.
Open Datasets Yes Los Angeles International Airport (LAX) Checkpoint Data from [Pita et al., 2008].
Dataset Splits No The paper does not explicitly provide training, validation, or test dataset splits with percentages or sample counts.
Hardware Specification No The paper states 'All algorithms run efficiently as expected (terminate within seconds using MATLAB)' but does not provide specific hardware details like CPU/GPU models, memory, or specific computing environments.
Software Dependencies No The paper mentions 'using MATLAB' but does not provide specific version numbers for MATLAB or any other software dependencies, libraries, or solvers.
Experiment Setup Yes All generated games have 20 targets and 10 resources. The reward ri (cost ci) of each target i is chosen uniformly at random from the interval [0, 10] ([ 10, 0]). ... For probabilistic information leakage, we randomly generate the probabilities that each target leaks information with the constraint Pn i=1 pi = 1 p0. For the case of leakage from small support (for simulated payoffs only), we randomly choose a support of size 5.