Causality

Spring semester 2017

General information

Lecturer Marloes Maathuis
Assistant Christina Heinze-Deml
Lectures Thu 8-10 HG D 7.1 >>
Course catalogue data >>

Course content

In statistics, we often consider prediction or classification problems. Many interesting questions in the sciences, however, do not fall within this framework. In this course we will consider questions of a causal nature, that is, questions about the causal mechanism behind a system or about predicting a system's behavior under manipulations. We will study concepts, theory and applications of modern causal inference methods.

Announcements

  • August 15th 2017:
    The question hour on August 21st (9:30 - 11:00 am) will take place in HG G 19.1.
  • July 16th 2017:
    We will hold another question hour on August 21st, please find the details here.
  • May 23rd 2017:
    We will hold a question hour on August 14th, please find the details here.
  • May 16th 2017:
    We fixed a typo in series 3 (exercise 3).
  • May 15th 2017:
    Please note: The pcalg package depends on the packages ‘graph’, ‘RBGL’, and ‘Rgraphviz’, which need to be installed from Bioconductor.
  • May 10th 2017:
    • No class on May 18th and May 25th.
    • The last series will be posted on May 12th and will be due on May 19th.
  • May 9th 2017:
    The oral exams will take place in the period August 21 - September 1.
  • May 4th 2017:
    Updated slides and R code for week 10; added slides for week 8 and R code for week 11.
  • March 30th 2017:
    No class on April 6th and April 20th.
  • March 23rd 2017:
    Updated slides on covariate adjustment (week 4).
  • March 16th 2017:
    Homework 1 is due on Monday, March 20th 2017. Please email your solutions to Christina Heinze-Deml or place them in the corresponding tray in HG J 68.
    The submission of solutions is only mandatory for PhD students who need ETH credit points. Students who need ECTS credit points have to take the exam.
  • March 10th 2017:
    Updated slides on DAGs (week 2).
  • February 8th 2017:
    Beginning of lecture: Thursday, 23/02/2017.

Course materials

Week Topic
Week 1 Introduction
Week 2 Graphical models
Week 3 Causal DAG models
Weeks 4 and 5 Causal DAG models and covariate adjustment
Week 6 Counterfactuals and instrumental variables
Week 7 No class
Week 8 Markov assumptions, faithfulness assumption, minimal I-map, perfect map, Markov equivalence, CPDAG
Week 9 No class
Week 10 Constraint-based structure learning
Week 11 Score-based structure learning
Week 12 LiNGAM

Exercises

Please email your solutions to Christina Heinze-Deml or place them in the corresponding tray in HG J 68. The submission of solutions is only mandatory for PhD students who need ETH credit points. Students who need ECTS credit points have to take the exam.

Exercises Solutions Due date
Series 1 Solutions 1 March 20th 2017
Series 2 Solutions 2 April 13th 2017
Series 3, data set: structure-learning.rda.
Please note: The pcalg package depends on the packages ‘graph’, ‘RBGL’, and ‘Rgraphviz’, which need to be installed from Bioconductor.		 Solutions 3 May 19th 2017

Literature

  • Freedman, Pisani and Purves (2007). Statistics. Fourth edition. Chapters 1-2.
  • Shalizi. Advanced Data Analysis from an Elementary Point of View. Draft available here. Chapters 20-25.
  • Højsgaard, Edwards and Lauritzen (2012). Graphical Models with R.
  • Pearl (2009). Causal inference in statistics: An overview. Available here.
  • Spirtes, Glymour and Scheines (2000). Causation, Prediction and Search. MIT Press. Available here.
  • Pearl (2009). Causality: Models, Reasoning and Inference. Wiley.
  • Pearl, Glymour and Jewell (2016). Causal Inference in Statistics: A Primer. Wiley.
  • Peters, Janzing and Schölkopf (2017). Elements of Causality. MIT Press. Draft available here.
  • Koller and Friedman (2009). Probabilistic Graphical Models. MIT Press. Available here.