Machine Learning for Causal Inference from Observational Data (26th July 2023)

Agenda

• 9:30-10:30
  • Welcome and Introduction to the Course
  • Lecture: Causal Inference using Machine Learning
• 10:30-11:00
  • Short Break
• 11:00-12:30
  • Lecture: (continued)
• 12:30-13:30
  • Lunch Break
• 13:30-15:00
  • Practical Part 1: Guided Example
• 15:00-15:30
  • Short Break
• 15:30-17:00
  • Practical Part 2: Do It Yourself
  • Consolidation, Discussion and Next Steps

Practical Parts

You will need a Python environment to complete the programming parts. The easiest option is to use Google Colab as it is ready to use straightaway. Alternatively, you can set up a local Python environment and run the code locally on your computer. If this is your choice, the latest Anaconda is recommended. Simply install it and you are ready to go. Instructions will be provided during the course about how to use Google Colab/Anaconda.

Signing-up for Google Colab

1. Create a Google account if you do not have one already.
2. Go to https://colab.research.google.com/.
3. If you see a “Sign in” button in the top right corner of the screen, click it and sign in using your Google account. If you see your account’s profile picture instead, you are already signed in.
4. In the top right corner of the screen, there is also a “Connect” button. Click it. A successful connection will confirm you are logged in correctly.
5. Feel free to explore the default “Welcome to Colaboratory” notebook (the one opened by default when you visit the website). Execute some code cells and familiarise yourself with the environment. This step is entirely optional as we will cover this in the course.

Further resources

• Introduction to Causal Inference online course.
• Recent causal inference surveys:
  • https://doi.org/10.1145/3397269
  • http://arxiv.org/abs/2002.02770
• EconML.
• Python courses.
• Scikit-learn docs.
• Scientific Python ecosystem - SciPy.