On the Inducibility of Stackelberg Equilibrium for Security Games

Authors: Qingyu Guo, Jiarui Gan, Fei Fang, Long Tran-Thanh, Milind Tambe, Bo An2020-2028

AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate our solution concept and proposed algorithmic implementation with extensive experiments. All results are obtained on a platform with a 2.60 GHz dual-core CPU and 8.0 GB memory. All linear programs are solved using the existing solver CPLEX (version 12.4). The random instances are generated as follows: rewards and penalties are all integers randomly drawn from [0, 5] and [−5, 0] respectively. Each schedule is randomly generated covering a fixed number l of targets and each target is ensured to be covered by at least one schedule. The resources are all homogeneous, i.e., Sr = S for any r ∈ R. Unless otherwise specified, all results are averaged on 100 randomly generated instances.
Researcher Affiliation Academia 1School of Computer Science and Engineering, Nanyang Technological University, {qguo005, boan}@ntu.edu.sg 2Department of Computer Science, University of Oxford, jiarui.gan@cs.ox.ac.uk 3School of Computer Science, Carnegie Mellon University, feifang@cmu.edu 4Department of Electronics and Computer Science, University of Southampton, ltt08r@ecs.soton.ac.uk 5Center for Artificial Intelligence in Society, University of Southern California, tambe@usc.edu
Pseudocode No The paper describes algorithmic steps in paragraph form, such as in the 'Computing an ISE' section, but it does not present formal pseudocode or algorithm blocks (e.g., Algorithm 1, Pseudocode).
Open Source Code No The paper does not provide an explicit statement about releasing source code or a direct link to a code repository for their methodology.
Open Datasets No The random instances are generated as follows: rewards and penalties are all integers randomly drawn from [0, 5] and [−5, 0] respectively. Each schedule is randomly generated covering a fixed number l of targets and each target is ensured to be covered by at least one schedule. The resources are all homogeneous, i.e., Sr = S for any r ∈ R.
Dataset Splits No The paper describes how its random instances are generated and the parameters used for these generations, but it does not specify any training, validation, or test dataset splits.
Hardware Specification Yes All results are obtained on a platform with a 2.60 GHz dual-core CPU and 8.0 GB memory.
Software Dependencies Yes All linear programs are solved using the existing solver CPLEX (version 12.4).
Experiment Setup Yes The random instances are generated as follows: rewards and penalties are all integers randomly drawn from [0, 5] and [−5, 0] respectively. Each schedule is randomly generated covering a fixed number l of targets and each target is ensured to be covered by at least one schedule. The resources are all homogeneous, i.e., Sr = S for any r ∈ R. Unless otherwise specified, all results are averaged on 100 randomly generated instances. [...] The game instances are randomly generated with l = 5, |R| = 5, |T| ranges from 50 to 400 with step size of 50, and |S| = |T|/2. [...] 500 instances are randomly generated with 200 targets, 1 resource, |S| ∈ {16, 18, 20} and l ∈ {16, 18, 20}.