Regret Minimization in Behaviorally-Constrained Zero-Sum Games

Authors: Gabriele Farina, Christian Kroer, Tuomas Sandholm

ICML 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conducted experiments to investigate the practical performance of our perturbed-regret-minimization approach when used to instantiate the CFR and CFR+ algorithms for computing approximate EFPE in EFGs. We compare these algorithms to state-of-the-art Nash-equilibrium-finding algorithms...
Researcher Affiliation Academia Carnegie Mellon University, Pittsburgh PA 15213 USA. Correspondence to: Gabriele Farina <gfarina@cs.cmu.edu>, Christian Kroer <ckroer@cs.cmu.edu>, Tuomas Sandholm <sandholm@cs.cmu.edu>.
Pseudocode Yes Algorithm 1 RM+ algorithm for generalized normal-form games played over finitely-generated convex polytopes. and Algorithm 2 Regret minimization algorithm for perturbed extensive-form games.
Open Source Code No The paper does not contain any explicit statement or link indicating that the source code for the methodology described in this paper is publicly available.
Open Datasets Yes We conducted the experiments on Leduc hold em poker (Southey et al., 2005), a widely-used benchmark in the imperfect-information game-solving community.
Dataset Splits No The paper describes the game setup and perturbations (e.g., 'k={3,5}' and 'uniform perturbations p(I, a) = ξ for all information sets I and actions a A(I), for ξ {0.1, 0.05, 0.01, 0.005, 0.001}'), but does not provide specific training, validation, or test dataset splits (e.g., percentages or sample counts).
Hardware Specification No No specific hardware details (e.g., GPU/CPU models, memory) used for running the experiments were mentioned.
Software Dependencies No The paper refers to various algorithms and techniques, but does not provide specific software dependencies with version numbers (e.g., programming languages, libraries, or solvers with their versions) required for replication.
Experiment Setup Yes We test our approach on games subject to different uniform perturbations p(I, a) = ξ for all information sets I and actions a A(I), for ξ {0.1, 0.05, 0.01, 0.005, 0.001}.