A Pair-Approximation Method for Modelling the Dynamics of Multi-Agent Stochastic Games

Authors: Chen Chu, Zheng Yuan, Shuyue Hu, Chunjiang Mu, Zhen Wang

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We verify the descriptive power of our model (a partial differential equation) across various games through comparisons with agent-based simulation results. To illustrate our model, we consider different game configurations in our experiments and numerically solve the developed partial differential equation in those games.
Researcher Affiliation Academia 1 School of Statistics and Mathematics, Yunnan University of Finance and Economics 2 School of Artificial Intelligence, OPtics and Electro Nics (i OPEN), Northwestern Polytechnical University 3 Shanghai Artificial Intelligence Laboratory 4 School of Cybersecurity, Northwestern Polytechnical University
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code Yes More details about the derivation of Equation (12) are presented in our supplementary material1. 1https://github.com/Zheng-YZ/AAAI2023SM
Open Datasets No The paper describes generating synthetic data for simulations (e.g., 'initial Q-values of agents follow different Beta distributions') but does not provide access information for a publicly available or open dataset. It refers to game configurations and simulation parameters rather than external datasets.
Dataset Splits No The paper describes simulation setups with different initial conditions but does not specify training, validation, or test dataset splits, as it primarily relies on agent-based simulations rather than pre-partitioned datasets.
Hardware Specification No The paper does not provide specific hardware details such as GPU models, CPU models, or cloud computing specifications used for running the experiments.
Software Dependencies No The paper does not provide specific software dependencies or library versions needed to replicate the experiment.
Experiment Setup Yes For the agent-based simulations, we set the population size n = 1000, the learning rate α = 0.4, and the temperature τ = 2 (Unless otherwise specified, the parameters are set in the same way for subsequent experiments). ... In (b), we set Q0(a1) Beta(20, 80, -0.1, 1.2), Q0(a2) Beta(80, 20, -0.1, 1.2).