Causality is a fundamental notion in science and engineering,
and one of the fundamental problems in the field is how to find the causal structure or the underlying
causal model. For instance, one focus of this workshop is on causal discovery, i.e.,
how can we discover causal structure over a set of variables from observational data with automated procedures?
Another area of interest is on how a causal perspective may help understand and solve advanced
machine learning problems.
Recent years have seen impressive progress in theoretical and algorithmic
developments of causal discovery from various types of data (e.g., from i.i.d. data, under distribution
shifts or in nonstationary settings, under latent confounding or selection bias, or with missing data),
as well as in practical applications (such as in neuroscience, climate, biology, and epidemiology).
However, many practical issues, including confounding, large scale of the data, the presence of
measurement error, and complex causal mechanisms, are still to be properly addressed, to achieve
reliable causal discovery in practice.
Moreover, causality-inspired machine learning (in the context of transfer learning,
reinforcement learning, deep learning, etc.) leverages ideas from causality to improve generalization,
robustness, interpretability, and sample efficiency and is attracting more and more interests in
Machine Learning (ML) and Artificial Intelligence. Despite the benefit of the causal view in transfer
learning and reinforcement learning, some tasks in ML, such as dealing with adversarial attacks and
learning disentangled representations, are closely related to the causal view but are currently
underexplored, and cross-disciplinary efforts may facilitate the anticipated progress.
This workshop aims to provide a forum for discussion for researchers
and practitioners in machine learning, statistics, healthcare, and other disciplines to share their
recent research in causal discovery and to explore the possibility of interdisciplinary collaboration.
We also particularly encourage real applications, such as in neuroscience, biology, and climate science,
of causal discovery methods.
There are two tracks of submissions: paper track and dataset track.
For the paper track, we invite submissions on all topics of causal discovery and causality-inspired ML,
including but not limited to:
- Causal discovery in complex environments, e.g., in the presence of distribution shifts,
latent confounders, selection bias, cycles, measurement error, small samples, or missing data
- Efficient causal discovery in large-scale datasets
- Causal effect identification and estimation
- Real-world applications of causal discovery, e.g. in neuroscience, finance, climate, and biology
- Assessment of causal discovery methods and benchmark datasets
- Causal perspectives on the problem of generalizability, transportability, transfer learning, and life-long learning
- Causally-enriched reinforcement learning and active learning
- Disentanglement, representation learning, and developing safe AI from a causal perspective
Submitted papers should follow the requirements for NeurIPS 2020 submissions. The length of submissions
is flexible, but is limited to eight content pages, including all figures and tables; additional pages
containing only references are allowed. Please format your submission using the NeurIPS 2020 LaTeX style
file. If needed, authors may additionally submit a supplementary material. According to the workshop
guidance provided by the conference, “work that is presented at the main NeurIPS conference should not
appear in a workshop, including as part of an invited talk.”
For the dataset track, we invite submissions of datasets from various fields, e.g., neuroscience, biology,
finance, and climate, that are appropriate for evaluating the performance of causal discovery methods.
Submissions should include (1) the collected dataset (the file or a link is required) and (2) a description
of the dataset in PDF format (with NeurIPS 2020 LaTeX style file), limited to four pages. The description
should include the “ground truth” causal structure from domain knowledge or experiments and the testing
results of at least one causal discovery method.
All accepted papers and datasets will be available on the workshop website. At the end of your paper
submission, please indicate whether you would like an extended version of the submission to be
considered for publication in a journal special issue. According to the feedback
from authors, we will further decide whether to publish selected papers in proceedings or a journal
- Submission: October 14, 2020
- Notification: October 30, 2020
- Camera-ready and slides: November 14, 2020
- Video submission: November 14, 2020
- Workshop: December 11 or December 12th, 2020
Papers and datasets can be submitted through CMT: https://cmt3.research.microsoft.com/CDML2020/.
Authors must not include any identifying information of the authors
(names, affiliations, etc.) or links and self-references that may reveal the authors' identities.
The organizers aim to provide feedback from three reviewers per submission, which will assess the
submission based on relevance, novelty, and potential for impact. Reviewers are asked to assess the
submission (Reject/Borderline/Accept) as well as provide written feedback. There will be no additional
The workshop will consist of four main parts: pre-recorded invited talks with a live discussion,
pre-recorded contributed presentations with a live discussion, a virtual poster session with spotlight
talks, and, to conclude, a panel discussion. We will have seven invited talks, covering recent
developments in causal discovery and inference, the connection between causal modeling and machine
learning, and applications of causal analysis. Importantly, we will make all videos available on our
website one week before the workshop, so attendees in all time zones can watch them beforehand and
submit questions. To transfer the social and forum-like aspects of in-person workshops to our virtual
workshop, during the workshop day we will have a dedicated Slack channel for introductions and virtual
coffee breaks in which we will try to foster social interaction and discussion on Slack and Zoom.
We will select four papers among all accepted ones for 10-minute contributed talks. Each paper to be
presented in the poster session will have a 2-minute spotlight introduction, followed by a poster
exhibition during lunch break and after closing remarks. Finally, in the panel discussion, we will
ask our invited speakers to discuss questions from the audience that are collected beforehand.
The detailed schedule will come soon.
Area Chairs: AmirEmad Ghassami, Antti Hyttinen, Biwei Huang, Christopher Meek, Daniel Malinsky, Dominik Janzing, Frederick Eberhardt, Ilya Shpitser, Mingming Gong, Peter Spirtes, Ruichu Cai, Sanghack Lee, Sara Magliacane, Shohei Shimizu, Tom Claassen, Yangbo He
Program Committee Members: AmirEmad Ghassami, Amir-Hossein Karimi, Biwei Huang, Bryan Andrews, Daniel Malinsky, Debarun Bhattacharjya, Eli Sherman, Erich Kummerfeld, Feng Xie, Ignavier Ng, Jaron Lee, Joseph Ramsey, Julius Kügelgen, Junzhe Zhang, Karthikeyan Shanmugam, Kenney Ng, Kun Kuang, Lin Liu, Mingming Gong, Murat Kocaoglu, Petar Stojanov, Ran Wang, Razieh Nabi, Rohit Bhattacharya