This repository contains the code to reproduce the results of our paper "Collective Intelligence in Decision-Making with Non-Stationary Experts".
Use runner.py to generate the result files:
python runner.py seed
Plots can subsequently be generated through the plots.ipynb notebook.
If you'd like to plot Figure 7 (normalized model error against average reward), pass the --extra-figure flag to runner.py. This will significantly slow down the experiment.
python runner.py seed --extra-figure