You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Exploring Multi-Agent Competition Dynamics: A Simulation Study on Reinforcement Learning and Collusion
The code that produces all the relevant figures for the paper can be found in $\texttt{main.ipynb}$. The files $\texttt{threeplayers.py}$ and $\texttt{twoplayers.py}$ are python modules that implement the sequential oligopoly games with two and three agents, respectively. $\texttt{utility.py}$ is a module consisting of a loose collection of utilities used in certain computations such as the Edgeworth Cycle benchmarks. $\texttt{runtime.ipynb}$ is used for the runtime analysis presented in the paper.
To run the Jupyter Notebook ($\texttt{ipynb}$) files, you can use the Jupyter Notebook application from a distribution like Anaconda (https://www.anaconda.com/).