This repository contains the code for our paper:
Adalina: Adaptive Linear Approximation for the Shapley Value and Beyond, which can be found at https://openreview.net/forum?id=hMTCuqjCW8.
The test/ directory contains two scripts used to verify the correctness of our implementation.
Follow the steps below to generate the experimental results and reproduce the figures from the paper.
Use the -p flag followed by the number of available CPUs to parallelize the runs.
-
To generate the main experimental results:
python main.py -p <number_of_cpus> -
To generate the results used for Figure 2:
python main_vr.py -p <number_of_cpus>
Once the results are ready, the figures can be plotted using:
python plot_adaptive.py
python plot_comparison.py
python plot_ps.py
python plot_vr.py