Causal Discovery in the Presence of Missing Values for Neuropathic Pain Diagnosis
Abstract
The missing data issue is a common phenomenon in many applications such as healthcare. When applying causal discovery algorithms, such as PC, to a data set with missing values, not properly handling the missing data issue might introduce bias and lead to wrong causal relations. In this work, we identify the potential errors of simply applying PC to data sets with missing values. Further, we extend the constraint-based causal discovery method PC to handle binary data sets with missing values for the application Neuropathic Pain diagnosis.
Cite
Text
Tu et al. "Causal Discovery in the Presence of Missing Values for Neuropathic Pain Diagnosis." ICML 2020 Workshops: Artemiss, 2020.Markdown
[Tu et al. "Causal Discovery in the Presence of Missing Values for Neuropathic Pain Diagnosis." ICML 2020 Workshops: Artemiss, 2020.](https://mlanthology.org/icmlw/2020/tu2020icmlw-causal/)BibTeX
@inproceedings{tu2020icmlw-causal,
title = {{Causal Discovery in the Presence of Missing Values for Neuropathic Pain Diagnosis}},
author = {Tu, Ruibo and Zhang, Kun and Bertilson, Bo Christer and Glymour, Clark and Kjellström, Hedvig and Zhang, Cheng},
booktitle = {ICML 2020 Workshops: Artemiss},
year = {2020},
url = {https://mlanthology.org/icmlw/2020/tu2020icmlw-causal/}
}