Audit Games with Multiple Defender Resources

Authors: Jeremiah Blocki, Nicolas Christin, Anupam Datta, Ariel Procaccia, Arunesh Sinha

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

Reproducibility Variable Result LLM Response
Research Type Experimental In addition, we experimentally demonstrate that this transformation significantly speeds up computation of solutions for a class of audit games and security games. In this section, we empirically demonstrate the speedup gains from our optimization transformation for both audit games and security games.
Researcher Affiliation Academia 1Carnegie Mellon University, USA; {arielpro@cs., jblocki@cs., danupam@, nicolasc@}cmu.edu 2University of Southern California, USA; aruneshs@usc.edu
Pseudocode Yes Algorithm 1: CONSTRAINT FIND(T, R) and Algorithm 2: APX SOLVE(l, EQ(j))
Open Source Code No The paper states 'Code was written in Matlab using the built-in large scale interior point method implementation of linear programming' but does not provide a concrete link to or statement about the public availability of their source code for the methodology described.
Open Datasets No The paper states that 'utilities were generated randomly from the range [0, 1]' for their experiments, indicating synthetic data, and does not provide access information for a publicly available or open dataset.
Dataset Splits No The paper does not specify explicit training, validation, or test dataset splits. It mentions 'random utilities' and running experiments for '5 runs' but no detailed split information for reproducibility.
Hardware Specification Yes Our experiments were run on a desktop with quad core 3.2 GHz processor and 6GB RAM.
Software Dependencies No The paper states 'Code was written in Matlab using the built-in large scale interior point method implementation of linear programming.' However, it does not provide specific version numbers for Matlab or any particular libraries/solvers used.
Experiment Setup Yes We used the same problem inputs in which utilities were generated randomly from the range [0, 1], a was fixed to 0.01, x was discretized with interval size of 0.005.