Welcome to the Intro to Causality course.

There is an increasing interest in using machine learning to solve decision making problems via causal inference and the use of causal inference to improve the robustness of machine learning models. This course will pay particular attention to the emerging intersection of deep learning & causal inference.

<aside> 📢 Announcement 1 [Sept 15 2025]: The class will now move to MS 3278

</aside>

<aside> 📢 Announcement 2 [Sept 15 2025]: Assignment 1 (Problem Set 1) is released. See this link.

</aside>

<aside> 📢 Announcement 3 [Sept 16 2025]: TA’s office hours have been changed to Thursdays 5pm - 6pm on Zoom [https://utoronto.zoom.us/j/89092860233]

</aside>

<aside> 📢 Announcement 4 [Sept 17 2025, Crowdmark Issue]: There is an issue with crowdmark since access permissions changed last week; consequently some of you received emails through the old systems while others did not. For the moment, please hand-write/iPad write your answers in some legible format that lets you upload it to crowdmark. We will send out another announcement when the updated submission site is ready.

</aside>

<aside> 📢 Announcement 5 [Sept 17 2025]: Note the following changes to the schedule and consequently the course and how it will progress through the semester. These changes were in part due to the total number of students in the class (since the course cap was raised) and consequently the planning around how to fit all student/project presentations in a manner that still interleaved it with course content.

a] We have updated the schedule of the course to move the paper presentations forward by two weeks and LLM & amortization lectures moved back. This will ensure that we are not jumping from student presentations into final project presentations directly.

b] We will have suggested project ideas posted soon!

c] We now have a planning spreadsheet with three tabs for the latter half of the course. You can find it here as well as on the tabs for the paper presentation and project presentation.

In the spreadsheet, the first tab indicates the paper presentation schedule and the dates corresponding to when each team will present. The second tab is a list of papers that you may choose to present. The third tab lists the dates for when the corresponding teams will have their final project presentation.

Student TODOs:

1) List your names and team-mates names in Tab 1. Each team should have three members exactly. In exceptional circumstances, and assuming there is no student who does not need a team, we may permit a two member team. The reason the class is organized this way is because cannot accommodate more teams than 16 due to the time-constraints presented by the class.

2) In Tab 2, put your team name next to the paper you will select. Papers are selected on a first-come first-serve basis.

If you do not have a team as yet and would like to find folks to work with, please look to the final tab in the spreadsheet and list your name and email.

</aside>

<aside> 📢 Announcement 6 [Sept 21 2025]: We have released a list of suggested project ideas here.

</aside>

<aside> 📢 Announcement 7 [Oct 14 2025]: Assignment 3 (Problem Set 2) is released. See this link.

</aside>

<aside> 📢 Announcement 8 [Oct 24 2025]: The deadline for Assignment 3 (Problem Set 2) has been extended to Monday, November 3.

</aside>

Overview

This course will cover basic principles in observational causal inference and machine learning and how the two can be combined in practice to assist with decision making from real world, high-dimensional datasets. Lectures will provide foundational knowledge of these new methodologies & the course assignments will involve real world data and synthetic data analysis.

It will be a mixed course comprising lectures, assignments and projects. Students will learn the basic concepts, nomenclature, and results in causality, along with advanced material characterizing recent applications of causality in machine learning.

This course is designed as a graduate level course to introduce computer scientists to concepts in causal inference. The course prerequisites include a strong background in probabilities, statistics, and Bayesian networks (such as factorization, d-separation, colliders, and Markov properties).

Note: The class will not cover prerequisite material. If you are rusty, please brush up; the Readings tab provides materials for this.

Instructors

Rahul G. Krishnan & Vahid Belazadeh

📅 What’s happening this week?