Since Faithfulness Fails: The Performance Limits of Neural Causal Discovery
Abstract
Neural-based causal discovery methods have recently improved in terms of scalability and computational efficiency. However, there are still opportunities for improving their accuracy in uncovering causal structures. We argue that the key obstacle in unlocking this potential is the faithfulness assumption, commonly used by contemporary neural approaches. We show that this assumption, which is often not satisfied in real-world or synthetic datasets, limits the effectiveness of existing methods. We evaluate the impact of faithfulness violations both qualitatively and quantitatively and provide a unified evaluation framework to facilitate further research.
Cite
Text
Olko et al. "Since Faithfulness Fails: The Performance Limits of Neural Causal Discovery." NeurIPS 2024 Workshops: CRL, 2024.Markdown
[Olko et al. "Since Faithfulness Fails: The Performance Limits of Neural Causal Discovery." NeurIPS 2024 Workshops: CRL, 2024.](https://mlanthology.org/neuripsw/2024/olko2024neuripsw-since/)BibTeX
@inproceedings{olko2024neuripsw-since,
title = {{Since Faithfulness Fails: The Performance Limits of Neural Causal Discovery}},
author = {Olko, Mateusz and Gajewski, Mateusz and Wojciechowska, Joanna and Kuciński, Łukasz and Morzy, Mikołaj and Sankowski, Piotr and Miłoś, Piotr},
booktitle = {NeurIPS 2024 Workshops: CRL},
year = {2024},
url = {https://mlanthology.org/neuripsw/2024/olko2024neuripsw-since/}
}