Causal Interpretability for Machine Learning – Problems, Methods and Evaluation


Raha Moraffah, Mansooreh Karami, Ruocheng Guo, Adrienne Raglin, Huan Liu

SIGKDD Explorations Newsletter 22(1), pages 18–33, 16 pages, June 2020
Association for Computing Machinery (ACM), New York, NY, USA

Machine learning models have had discernible achievements in a myriad of applications. However, most of these models are black-boxes, and it is obscure how the decisions are made by them. This makes the models unreliable and untrustworthy. To provide insights into the decision making processes of these models, a variety of traditional interpretable models have been proposed. Moreover, to generate more humanfriendly explanations, recent work on interpretability tries to answer questions related to causality such as "Why does this model makes such decisions?" or "Was it a specific feature that caused the decision made by the model?". In this work, models that aim to answer causal questions are referred to as causal interpretable models. The existing surveys have covered concepts and methodologies of traditional interpretability. In this work, we present a comprehensive survey on causal interpretable models from the aspects of the problems and methods. In addition, this survey provides in-depth insights into the existing evaluation metrics for measuring interpretability, which can help practitioners understand for what scenarios each evaluation metric is suitable.

(keywords) counterfactuals, interpratablity, causal inference, explainability, machine learning
 @article{causalinterpretability-kddnews22,
address = {New York, NY, USA},
author = {Moraffah, Raha and Karami, Mansooreh and Guo, Ruocheng and Raglin, Adrienne and Liu, Huan},
doi = {10.1145/3400051.3400058},
issn = {1931-0145},
journal = {SIGKDD Explorations Newsletter},
keywords = {counterfactuals, interpratablity, causal inference, explainability, machine learning},
month = jun,
number = 1,
numpages = 16,
pages = {18--33},
publisher = {Association for Computing Machinery},
title = {Causal Interpretability for Machine Learning -- Problems, Methods and Evaluation},
url = {https://doi.org/10.1145/3400051.3400058},
volume = 22,
year = 2020

Publication

— authors

Raha Moraffah, Mansooreh Karami, Ruocheng Guo, Adrienne Raglin, Huan Liu

— status

published

— sort

article in journal

Venue

— journal

SIGKDD Explorations Newsletter

— volume

22

— issue

1

— pages

18–33

— publication date

June 2020

URLs

original page  |  open access PDF

Identifiers

— DOI

10.1145/3400051.3400058

— print ISSN

1931-0145

BibTeX

— BibTeX ID
causalinterpretability-kddnews22

Files

Open Access PDF

View this PDF full screen

You do not have the plugin required to display this PDF file. You can still download it: 4._CR._25._Causal_Explainability_Survey-final.pdf

Partita IVA: 01131710376 - Copyright © 2008-2022 APICe@DISI Research Group - PRIVACY