Objective

This course will focus on recent developments on causal graphical models, including causal reasoning and causal discovery. The objective is that the students know the fundamentals of causal graphical models and the main advances in reasoning and learning, so they can do research in this area.

Outline

Introduction to causal models
Causal Bayesian networks
Causal reasoning
Causal decision making
Causal discovery
Applications

Text Books

J. Pearl, Causality, Second Edition, Cambridge U. P., 2009

P. Spirtes, C- Glymour, R. Scheines, Causation, Prediction, and Search, Springer, 1993

J. Woodward, Making Things Happen: A theory of Causal Explanation, Oxford U. P., 2003

J. Peters, D. Janzing, B. Scholkpf, Elements of Causal Inference, MIT Press, 2017.

J. Pearl, D. Mackienze, The Book of Why, Basic Books, 2018.

Homework

Week 1/2: REPRESENTATION - Read and prepare a presentation of the following papers to be discussed in class:

Causal Models, Stanford Encyclopedia of Philosophy
An Introduction to Causal Inference, Richard Scheines
Introduction to Causal Inference, Spirtes, JMLR 11, 2010
Pearl, Causality, 2nd Edition - Chapter 1: 1.3 -- 1.5

Week 3: REASONING - Read and prepare a presentation of the following papers to be discussed in class:

Elementos of Causal Reasoning - Chapter 2: Assumptions for Causal Inference
Pearl, Causality, 2nd Edition - Chapter 3: Causal Diagrams and the Identification of Causal Effects

Week 4: DECISIONS - Read and prepare a presentation of the following papers to be discussed in class:

Introduction to Decision Theory and its extension to Causal Decision Theory (Mauricio)
Pearl, Causality, 2nd Edition - Chapter 4: Actions, Plans and Direct Effects
S. J. Gershman. Reinforcement learning and causal models. The Oxford handbook of causal reasoning, page 295, 2017
C. Lu, B. Scholkopf, and J. M. Hernández-Lobato. Deconfounding reinforcement learning in observational settings. CoRR, abs/1812.10576, 2018.

Week 5: TIME - Read and prepare a presentation of the following papers to be discussed in class:

D. Malinsky, P. Spirtes, Causal Structure Learning from Multivariate Time Series in Settings with Unmeasured Confounding, Proceedings of 2018 ACM SIGKDD Workshop on Causal Disocvery
A. Barnard, D. Page, Causal Structure Learning via Temporal Markov Networks, PGM 2018
J. Runge, Causal network reconstruction from time series: From theoretical assumptions to practical estimation

Week 6/7: DISCOVERY - Read and prepare a presentation of the following papers to be discussed in class:

Pearl, Causality, 2nd Edition - Chapter 2: A Theory of Inferred Causation
C. Glymour, K. Zhang, P. Spirtes, Review of Causal Discovery Methods Based on Graphical Models, Frontiers in Genetics, June 2019
A. Hauser, P. Buhlmann, Two optimal strategies for active learning of causal models from interventions, PGM 2012 (in IJAR 2014)
D. Eaton, K. Murphy, Exact Bayesian structure learning from uncertain interventions, Artificial Intelligence and Statistics, 2007
S. Shimizu, P. Hoyer, A. Hyvarinen, A. Kirmenen, A linear-non-Gaussian acyclic model for causal discovery, JMLR 7, 2006

Week 8: Knowledge Transfer

J. Pearl, E. Barienboim, External validity: from Do-calculus to transportability across populations, Statistical Science, Vol. 29, 2014.

9: FINAL PROJECT

Project proposal (one page) - July 2nd
Project report (paper in Lecture Notes style, max. 10 pages) - July 16
Project presentation (max. half hour) - July 16

NOTE: It is expected that the presentation of papers is not limited to just repeating the contents, but making a synthesis and analysis, and enriching the presentation with additional resources and dicussion.

FINAL PROJECT

At the end of the course it is expected that each student presents a final project that could be:

A review of some specific area of causal graphical models
An implementation of an algorithm for inference or learning with causal models

A short paper describing the project and a presentation in class are requiered.

EVALUATION

Presentations and discussions in class - 70%

Final Project - 30%

PRESENTATIONS

Causal Models - Stanford Encyclopedia of Philosophy
Introduction to causal models, Spirtes
An introduction to causal models, Scheine
Elements of causal inference, Peters et al.

Causal Bayesian Networks, Sec. 1.3, Pearl
Causal Bayesian Networks, Sec. 1.4, Pearl
Terminology - Sec. 1.5, Pearl
Theory of inferred causation, Ch. 2, Pearl
Theory of inferred causation, Sec. 2.6-9, Pearl
Interventions in Markovian Models, Sec. 3.2, Pearl
Controlling confounding bias, Sec. 3.3, Pearl
A Calculus of Intervention, Sec. 3.4, Pearl
Action, Plans and Direct effect, Sec. 4.2-3, Pearl
Identification of Dynamic Plans, Sec. 4.4, Pearl
Identification of Dynamic Plans, Sec. 4.5, Pearl
Direct vs Total Effects, Sec. 4.5, Pearl

Non Gaussian or Non Linear Methods,
Review of causal discovery methods, Glymour, Zhang, Spirtes
Causal structure learning from time series, Malinsky & Srites
Causal structure learning via temporal Markov networks, Barnard & Page
LiNGAM, Shimizu, et al.
Exact Bayesian structure learning from uncertain interventions, Eaton & Murphy

Videos

Machine learning expert Jonas Peters of the University of Copenhagen presents “Four Lectures on Causality”.

Part 1: https://www.youtube.com/watch?v=zvrcyqcN9Wo&list=PLlMMtlgw6qNjROoMNTBQjAcdx53kV50cS

Part 2: https://www.youtube.com/watch?v=bHOGP5o3Vu0&list=PLlMMtlgw6qNjROoMNTBQjAcdx53kV50cS

Part 3: https://www.youtube.com/watch?v=Jp4UcgpVA2I&list=PLlMMtlgw6qNjROoMNTBQjAcdx53kV50cS

Part 4: https://www.youtube.com/watch?v=ytnr_2dyyMU&list=PLlMMtlgw6qNjROoMNTBQjAcdx53kV50cS

References

Spirtes, P. (2010).
Introduction to Causal Inference.
Journal of Machine Learning Research, 11(May), 1643–1662.
Retrieved from http://www.jmlr.org/papers/v11/spirtes10a.html

Kalisch, M., & Bühlmann, P. (2014).
Causal Structure Learning and Inference: A Selective Review.
Quality Technology & Quantitative Management, 11(1), 3–21.
https://doi.org/10.1080/16843703.2014.11673322

Zhou, Y. (2007).
Structure Learning of Probabilistic Graphical Models: A Comprehensive Survey.
Retrieved from http://arxiv.org/abs/1111.6925

Zhang, K., Schölkopf, B., Spirtes, P., & Glymour, C. (2018).
Learning causality and causality-related learning: Some recent progress.
National Science Review.
https://doi.org/10.1093/nsr/nwx137

Malinsky, D., & Danks, D. (2018).
Causal discovery algorithms: A practical guide.
Philosophy Compass, 13(1), e12470.
https://doi.org/10.1111/phc3.12470

Maathuis, M. H., & Nandy, P. (2015).
A review of some recent advances in causal inference.
Handbook of Big Data, 387–407.
Retrieved from http://arxiv.org/abs/1506.07669

Drton, M., & Maathuis, M. H. (2017).
Structure Learning in Graphical Modeling.
SSRN (Vol. 4). Annual Reviews.
https://doi.org/10.1146/annurev-statistics-060116-053803

Ellis, B., & Wong, W. H. (2008).
Learning Causal Bayesian Network Structures From Experimental Data.
Journal of the American Statistical Association, 103(June 2013), 778–789.
https://doi.org/10.1198/016214508000000193

Zuk, O., Margel, S., & Domany, E. (2012).
On the number of samples needed to learn the correct structure of a Bayesian network.
ArXiv Preprint ArXiv:1206.6862.
Retrieved from http://arxiv.org/abs/1206.6862

Dodge, Y., & Rousson, V. (2001).
On Asymmetric Properties of the Correlation Coefficient in the Regression Setting.
The American Statistician, 55(1), 51–54.
https://doi.org/10.1198/000313001300339932doi.org/10.1198/000313001300339932

Pearl, "Causal and Counterfactual Inference,"Forthcoming section in The Handbook of Rationality, MIT press.
      https://ucla.in/2Iz9myt

Pearl, "A note on oxygen, matches and fires," On Non-manipulable Causes," September 2018.
      https://ucla.in/2Qb1h6v

Pearl, "Does Obesity Shorten Life? Or is it the Soda? On Non-manipulable Causes," https://ucla.in/2EpxcNU
      Journal of Causal Inference, 6(2), online, September 2018.

Pearl, "The Seven Tools of Causal Inference with Reflections on Machine Learning," July 2018
      https://ucla.in/2umzd65 Forthcoming, Communications of ACM.

Cinelli and Pearl, "On the utility of causal diagrams in modeling attrition: a practical example," April 2018.
      https://ucla.in/2L8KAWw Forthcoming, Journal of Epidemiology.

Pearl and Bareinboim, "A note on `Generalizability of Study Results'," April 2018. Forthcoming, Journal of Epidemiology.
https://ucla.in/2NIsI6B

Causal effects

Hyttinen, A., Eberhardt, F., & Matti, J. (2015).
Do-calculus when the True Graph Is Unknown.
In Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence (pp. 395–404).

Maathuis, M. H., Kalisch, M., & Bühlmann, P. (2009).
Estimating high-dimensional intervention effects from observational data.
Annals of Statistics, 37(6 A), 3133–3164.
https://doi.org/10.1214/09-AOS685

Nandy, P., Maathuis, M., & Richardson, T. (2014).
Estimating the effect of joint interventions from observational data in sparse high-dimensional settings.
ArXiv Preprint ArXiv:1407.2451, 1–46.
Retrieved from http://arxiv.org/abs/1407.2451

Causality + RL

I. Dasgupta, J. X. Wang, S. Chiappa, J. Mitrovic, P. A. Ortega, D. Raposo, E. Hughes, P. Battaglia, M. Botvinick, and Z. Kurth-Nelson. Causal reasoning from meta-reinforcement learning. CoRR, abs/1901.08162, 2019.

S. J. Gershman. Reinforcement learning and causal models. The Oxford handbook of causal reasoning, page 295, 2017.

S. Ho. Causal learning versus reinforcement learning for knowledge learning and problem solving. In The Workshops of the The Thirty-First AAAI Conference on Artificial Intelligence, Saturday, February 4-9, 2017, San Francisco, California, USA, volume WS-17 of AAAI Workshops. AAAI Press, 2017.

C. Lu, B. Scholkopf, and J. M. Hernández-Lobato. Deconfounding reinforcement learning in observational settings. CoRR, abs/1812.10576, 2018.

C. Yu, Y. Dong, J. Liu, and G. Ren. Incorporating causal factors into reinforcement learning for
dynamic treatment regimes in HIV. BMC Med. Inf. & Decision Making, 19-S(2):19–29, 2019.

Tools & Data

Causality Benchmark Platform

Other external links

Conferences:

Uncertainty in AI

Probabilistic Graphical Models:

Projects

Resumen [PDF] [PPTX]