Designing the Game to Play: Optimizing Payoff Structure in Security Games

Authors: Zheyuan Ryan Shi, Ziye Tang, Long Tran-Thanh, Rohit Singh, Fei Fang

IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Numerical evaluation: We provide extensive experimental evaluation for the proposed algorithms. For problems with L1-norm form budget constraint, we show that the branchand-bound approach with an additive approximation guaran- Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI-18) tee can solve up to hundreds of targets in a few minutes.
Researcher Affiliation Academia 1 Swarthmore College, USA 2 Carnegie Mellon University, USA 3 University of Southampton, UK 4 World Wide Fund for Nature, Cambodia
Pseudocode Yes Algorithm 1 Branch-and-bound [...] Algorithm 2 PTAS for a special case in L1 [...] Algorithm 3 Algorithm for budget in L0-norm form
Open Source Code No The paper does not provide concrete access to source code for the methodology described. A footnote provides a link to the arXiv version of the paper itself: "https://arxiv.org/abs/1805.01987".
Open Datasets No The original payoff structures are randomly generated integers between 1 and 2n with penalties obtained by negation (recall n is the number of targets). Budget and weights of the manipulations are randomly generated integers between 1 and 4n.
Dataset Splits No The paper does not provide specific dataset split information (exact percentages, sample counts, or detailed splitting methodology).
Hardware Specification Yes For each problem size, we run 60 experiments on a PC with Intel Core i7 processor.
Software Dependencies No Gurobi is used for solving MILPs, which is terminated when either time limit (15 min) or optimality gap (1%) is achieved. (A version number for Gurobi is not provided).
Experiment Setup Yes Gurobi is used for solving MILPs, which is terminated when either time limit (15 min) or optimality gap (1%) is achieved. [...] We set ρ0 = maxi T Ra i 4(Rd i P d i ) which gives an additive 1 2-approximate solution.