NeurIPS 2018 Workshop on Causal Learning
Friday 7th, December 2018 @ Room 220 C, Palais des Congrès de Montréal
The route from machine learning to artificial intelligence remains uncharted. Recent efforts describe some of the conceptual problems that lie along this route [4, 9, 12]. The goal of this workshop is to investigate how much progress is possible by framing these problems beyond learning correlations, that is, by uncovering and leveraging causal relations:
- Machine learning algorithms solve statistical problems (e.g. maximum likelihood) as a proxy to solve tasks of interest (e.g. recognizing objects). Unfortunately, spurious correlations and biases are often easier to learn than the task itself , leading to unreliable or unfair predictions. This phenomenon can be framed as causal confounding.
- Machines trained on large pools of i.i.d. data often crash confidently when deployed in different circumstances (e.g., adversarial examples, dataset biases ). In contrast, humans seek prediction rules robust across multiple conditions. Allowing machines to learn robust rules from multiple environments can be framed as searching for causal invariances [2, 11, 16, 17].
- Humans benefit from discrete structures to reason. Such structures seem less useful to learning machines. For instance, neural machine translation systems outperform those that model language structure. However, the purpose of this structure might not be modeling common sentences, but to help us formulate new ones. Modeling new potential sentences rather than observed ones is a form of counterfactual reasoning [8, 9].
- Intelligent agents do not only observe, but also shape the world with actions. Maintaining plausible causal models of the world allows to build intuitions, as well as to design intelligent experiments and interventions to test them [16, 17]. Is causal understanding necessary for efficient reinforcement learning?
- Humans learn compositionally; after learning simple skills, we are able to recombine them quickly to solve new tasks. Such abilities have so far eluded our machine learning systems. Causal models are compositional, so they might offer a solution to this puzzle .
- Finally, humans are able to digest large amounts of unsupervised signals into a causal model of the world. Humans can learn causal affordances, that is, imagining how to manipulate new objects to achieve goals, and the outcome of doing so. Humans rely on a simple blueprint for a complex world: models that contain the correct causal structures, but ignore irrelevant details [16, 17].
We cannot address these problems by simply performing inference on known causal graphs. We need to learn from data to discover plausible causal models, and to construct predictors that are robust to distributional shifts. Furthermore, much prior work has focused on estimating explicit causal structures from data, but these methods are often unscalable, rely on untestable assumptions like faithfulness or acyclicity, and are difficult to incorporate into high-dimensional, complex and nonlinear machine learning pipelines. Instead of considering the task of estimating causal graphs as their final goal, learning machines may use notions from causation indirectly to ignore biases, generalize across distributions, leverage structure to reason, design efficient interventions, benefit from compositionality, and build causal models of the world in an unsupervised way.
Call for papers
Submit your anonymous, NeurIPS-formatted manuscript here. All accepted submissions will require a poster presentation. A selection of submissions will be awarded a 5-minute spotlight presentation. For each of the top 20 papers, we can provide the option to purchase 1 NeurIPS workshop ticket. We welcome conceptual, thought-provoking material, as well as research agendas, open problems, new tasks, and datasets.
28 October 2018 November 2, 23:59 PM GMT-3
Acceptance notifications: 12 November 2018
Each talk contains 20 minutes of presentation and 10 minutes for discussion. Each discussion will be prepared and chaired by one of the workshop organizers, with the intention to promote an interactive and engaging conversation between the speaker and the audience.
- 09:00-09:15: Check-in, set up of contributed posters
- 09:15-09:30: Opening remarks, David Lopez-Paz
- 09:30-10:00: Invited talk, Bernhard Schölkopf: Learning Independent Mechanisms
- 10:00-10:30: Invited talk, Pietro Perona: From micro-variables to macro-causes
- 10:30-11:00: Coffee break and poster discussions
- 11:00-11:30: Invited talk, Elias Bareinboim: On Causal Reinforcement Learning
- 11:30-12:00: Invited talk, Isabelle Guyon: Evaluating causation coefficients
- 12:00-12:30: Invited talk, David Blei: The blessings of multiple causes
- 12:30-14:00: Lunch break
- 14:00-14:30: Invited talk, Nicolai Meinshausen: Causality and distributional robustness
- 14:30-15:00: Invited talk, Csaba Szepesvári: Model-free vs. model-based learning in a causal world: Some stories from online learning to rank
- 15:00-15:30: Coffee break and poster discussions
- 15:30-17:00: Panel Discussion with Elias Bareinboim, Isabelle Guyon, David Blei, Nicolai Meinshausen, Csaba Szepesvári, Sara Magliacane and Yoshua Bengio: Datasets and benchmarks for causal learning
- 17:00-18:00: Contributed spotlights and poster discussions.
- 18:00- : Additional poster discussions.
- Lars Buesing, Theophane Weber, Yori Zwols, Nicolas Heess, Sebastien Racaniere, Arthur Guez and Jean-Baptiste Lespiau: Woulda, Coulda, Shoulda: Counterfactually-Guided Policy Search
- Sonali Parbhoo, Mario Wieser and Volker Roth: Cause-Effect Deep Information Bottleneck For Incomplete Covariates
- Ricardo Pio Monti, Kun Zhang and Aapo Hyvarinen: NonSENS: Non-Linear SEM Estimation using Non-Stationarity
- Pablo Rivas and Aishwarya Pagalla: Rule-Based Sentence Quality Modeling and Assessment using Deep LSTM Features
- Amanda Gentzel, Dan Garant and David Jensen: The Case for Evaluating Causal Models Using Interventional Measures and Empirical Data
- Tengyang Xie, Yu-Xiang Wang and Yifei Ma: Marginalized Off-Policy Evaluation for Reinforcement Learning
- Sara Magliacane, Thijs van Ommen, Tom Claassen, Stephan Bongers, Philip Versteeg and Joris Mooij: Domain Adaptation by Using Causal Inference to Predict Invariant Conditional Distributions
- Adarsh Subbaswamy, Peter Schulam and Suchi Saria: Learning Predictive Models That Transport
- Pim de Haan, Dinesh Jayaraman and Sergey Levine: Causal Confusion in Imitation Learning
- Dan Roberts and Max Kleiman-Weiner: Causality in Physics and Effective Theories of Agency
- Angelina Wang, Thanard Kurutach, Aviv Tamar and Pieter Abbeel: Learning Robotic Manipulation through Visual Planning and Acting
- Jiman Kim, Dongha Bahn, Chanjong Park, Jaeil Jung, Junik Jang and Jonghee Hong: Real Defect Image Classification through Hierarchical Data Augmentation
- Navin Goyal, Anand Louis and Karthik Abinav Sankararaman: Stability of Linear Structural Equation Model of Causal Inference
- Alexander Marx and Jilles Vreeken: Stochastic Complexity for Testing Conditional Independence on Discrete Data
- Eustache Diemert, Amélie Héliou and Christophe Renaudin: Off-policy learning for causal advertising
- Kailash Budhathoki, Mario Boley and Jilles Vreeken: Rule Discovery for Exploratory Causal Reasoning
- David Rohde: A Bayesian Solution to the M-Bias Problem
- Gabriel Schamberg and Todd P. Coleman: Quantifying Context-Dependent Causal Influences
- Jeric Briones, Takatomi Kubo and Kazushi Ikeda: Detecting switching causal interactions using hierarchical segmentation approach
- Xin Guo, Anran Hu, Renyuan Xu and Junzi Zhang: Consistency and Computation for Regularized Maximum Likelihood Estimation of Multivariate Hawkes Processes
- Andrew Melnik, Sascha Fleer, Malte Schilling and Helge Ritter: Modularization of End-to-End Learning: Case Study in Arcade Games
- Lequn Wang, Yi Su, Michele Santacatterina and Thorsten Joachims: CAB: Continuous Adaptive Blending Estimator for Policy Evaluation and Learning
- Murat Kocaoglu, Sanjay Shakkottai, Alexandros G. Dimakis, Constantine Caramanis and Sriram Vishwanath: Entropic Latent Variable Discovery
- Zhichong Fang, Aman Agarwal and Thorsten Joachims: Intervention Harvesting for Context-Dependent Examination-Bias Estimation
- Debmalya Mandal: Weighted Tensor Completion for Time-Series Causal Inference
- Krzysztof Chalupka, Pietro Perona, Frederick Eberhardt (2015): Visual Causal Feature Learning
- Christina Heinze-Deml, Nicolai Meinshausen (2018): Conditional Variance Penalties and Domain Shift Robustness
- Fredrik D. Johansson, Uri Shalit, David Sontag (2016): Learning Representations for Counterfactual Inference
- Brenden Lake (2014): Towards more human-like concept learning in machines: compositionality, causality, and learning-to-learn
- Brenden M. Lake, Tomer D. Ullman, Joshua B. Tenenbaum, Samuel J. Gershman (2016): Building Machines That Learn and Think Like People
- David Lopez-Paz, Krikamol Muandet, Bernhard Schölkopf, Ilya Tolstikhin (2015): Towards a Learning Theory of Cause-Effect Inference
- David Lopez-Paz, Robert Nishihara, Soumith Chintala, Bernhard Schölkopf, Léon Bottou (2017): Discovering Causal Signals in Images
- Judea Pearl (2009): Causality: Models, Reasoning, and Inference
- Judea Pearl (2018): The Seven Pillars of Causal Reasoning with Reflections on Machine Learning
- Jonas Peters, Joris Mooij, Dominik Janzing, Bernhard Schölkopf (2014): Causal Discovery with Continuous Additive Noise Models
- Jonas Peters, Peter Bühlmann, Nicolai Meinshausen (2016): Causal inference using invariant prediction: identification and confidence intervals
- Jonas Peters, Dominik Janzing, Bernhard Schölkopf (2017): Elements of Causal Inference: Foundations and Learning Algorithms
- Peter Spirtes, Clark Glymour, Richard Scheines (2001): Causation, Prediction, and Search
- Bob L. Sturm (2016): The HORSE conferences
- Dustin Tran, David M. Blei (2017): Implicit Causal Models for Genome-wide Association Studies
- Michael Waldmann (2017): The Oxford Handbook of Causal Reasoning
- James Woodward (2005): Making Things Happen: A Theory of Causal Explanation
- Antonio Torralba, Alyosha Efros (2011): Unbiased look at dataset bias