Regret-Based Optimization and Preference Elicitation for Stackelberg Security Games with Uncertainty

Authors: Thanh Nguyen, Amulya Yadav, Bo An, Milind Tambe, Craig Boutilier

AAAI 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results validate the effectiveness of our approaches.
Researcher Affiliation Academia 1University of Southern California, Los Angeles, CA 90089 {thanhhng, amulyaya, tambe}@usc.edu 2Nanyang Technological University, Singapore 639798 boan@ntu.edu.sg 3University of Toronto, Canada M5S 3H5 cebly@cs.toronto.edu
Pseudocode Yes Algorithm 1: Constraint-generation (MIRAGE)
Open Source Code No The paper does not provide concrete access to its source code, nor does it explicitly state that the code is available.
Open Datasets No The paper states that games were 'generated using GAMUT' and describes the generation process, but does not refer to or provide access information for a publicly available or open dataset in the traditional sense of a pre-existing collection of data.
Dataset Splits No The paper describes how game instances were randomly generated and evaluated, but it does not specify explicit training, validation, or test dataset splits.
Hardware Specification Yes All experiments were run on a 2.83GHz Intel processor with 4GB of RAM
Software Dependencies Yes using CPLEX 12.3 for LP/MILPs and KNITRO 8.0.0.z for nonlinear optimization.
Experiment Setup Yes Upper and lower bounds for payoff intervals are generated randomly from [ 14, 1] for penalties and [1, 14] for rewards, with the difference between the upper and lower bound (i.e., interval size) exactly 2 (this gives payoff uncertainty of roughly 30%). All results are averaged over 120 instances (20 games per covariance value) and use eight defender resources unless otherwise specified.