Undergraduate: Undergraduate causal inference and machine learning
Use causal models to define the effect of an action. Separate identification from estimation. Study graphs, experiments, hidden causes, discovery, and counterfactual reasoning before connecting these ideas to quantum systems.
Lessons in this level¶
| Lesson | Objective | Lab |
|---|---|---|
| L4-01 · Structural causal models | Replace a structural equation to calculate an intervention and compare it with observation. | Confounding and intervention |
| L4-02 · DAGs and d-separation | Use path rules to decide which conditional independences a causal graph implies. | Confounding and intervention |
| L4-03 · Confounders, mediators, and colliders | Choose adjustment variables by their graph role instead of adjusting for every available variable. | Confounding and intervention |
| L4-04 · Backdoor adjustment | Standardize outcome risks across a valid adjustment set to calculate an intervention effect. | Confounding and intervention |
| L4-05 · Frontdoor identification | Calculate a frontdoor effect and state all three graphical conditions that justify it. | Confounding and intervention |
| L4-06 · Potential outcomes and randomization | Define an average treatment effect and identify the assumptions that connect it to an observed contrast. | Confounding and intervention |
| L4-07 · Identification before estimation | Show how two models can fit the same observations and disagree about an intervention. | Confounding and intervention |
| L4-08 · Causal discovery and its limits | Explain Markov equivalence, faithfulness, and the limits of discovering directions from observational data. | Confounding and intervention |
| L4-09 · Counterfactual reasoning | Use abduction, action, and prediction to answer a counterfactual in an explicit structural model. | Confounding and intervention |
| L4-10 · Causal machine learning, transport, and fairness | Define a causal learning target and explain the additional assumptions needed for transport and fairness claims. | Confounding and intervention |
Watch & listen¶
4 - Causal Models. Brady Neal - Causal Inference · 48 min · 2020. Brady Neal’s lecture on causal models.
5.4 - Pearl’s do-calculus: Brady Neal - Causal Inference- Introduction to Causal Graphical Models: Graphs, d-separation, do-calculus: Simons Institute for the Theory of Computing
- Classical Strategies and Concepts in Causal Discovery: Simons Institute for the Theory of Computing
- Keynote: Judea Pearl - The New Science of Cause and Effect: PyData
- Causal Representation Learning: A Natural Fit for Mechanistic Interpretability: Mila - Institut québécois d’IA
- Causality, part 1 - Bernhard Schölkopf - MLSS 2020, Tübingen: virtual mlss2020
Sean Carroll’s Mindscape: Science, Society, Philosophy, Culture, Arts, and Ideas: Judea Pearl on Cause and Effect (2022-05-09), 77 min. Sean Carroll talks with Judea Pearl about cause and effect.- Lex Fridman Podcast: Judea Pearl: Causal Reasoning, Counterfactuals, Bayesian Networks, and the Path to AGI (2019-12-11), 83 min. Lex Fridman interviews Judea Pearl on causal reasoning, counterfactuals and Bayesian networks.
- The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence): Weakly Supervised Causal Representation Learning with Johann Brehmer (2022-12-15), 47 min. TWIML episode with Johann Brehmer on weakly supervised causal representation learning.
- Data Skeptic: Graphs for Causal AI (2025-05-24), 41 min. Data Skeptic episode on graphs for causal AI.
- Decoding Causality (the course author’s podcast): S1E04. Untangling the Web: How to Overcome Hidden Bias in Data. Confounding and how experiments, causal diagrams, and adjustment remove it; pairs with L4-03 and L4-04.
- Decoding Causality (the course author’s podcast): S1E05. Breaking the Illusion: How to Prove Causation Without Experiments. How to argue for causation when an experiment is impossible or unethical.
- Decoding Causality (the course author’s podcast): S1E07. Climbing the Causal Mountain: How to Predict the Impact of Interventions. Predicting the effect of an intervention, the second rung of the ladder of causation; pairs with L4-05.
- Decoding Causality (the course author’s podcast): S1E08. Imagining the Impossible: How Counterfactuals Shape Our World. Counterfactuals with structural causal models and potential outcomes; pairs with L4-09.
- Decoding Causality (the course author’s podcast): S1E09. Tracing the Invisible: How Causes Travel Through the World. Mediation: how causes travel to their effects, with path diagrams.
- Decoding Causality (the course author’s podcast): Closing the Loop: What We Learned About Causality. A closing recap that ties the ladder of causation, counterfactuals, mediation, and bias together.
More for every level on the Watch and listen page.
- Bareinboim, E. (2025). Causal Artificial Intelligence: A Roadmap for Building Causally Intelligent Systems. https://causalai-book.net/