CIML-R: Causally Informed Machine Learning Based on Feature Relevance

This is a minimal example for the method and evaluation presented in the paper CIML-R: Causally Informed Machine Learning Based on Feature Relevance.

Getting started

Install dependencies

> apt install -y graphviz 
> conda create -n cimlr python=3.10 pip 
> conda activate cimlr
> pip install -r requirements.txt

Start a jupyter notebook

> jupyter notebook

Then, you can use the notebook cimlr-evaluation.ipynb to run the experiment.