Causal Inference for Health Data
Winter 2026

BruinLearn

Logistics

Time TR 11am - 12.15pm
Location MathSci 5200
Instructor Drago Plecko (OH: Thu, 1-2pm, MathSci 8105E)
Syllabus CIHD Syllabus (pdf)
Course Code Repository Github Repo

Lectures

Week 1
Lecture 1   Introduction
Jan 6
  • - Logistics; Motivation; Causal Data Science; Pearl's Causal Hierarchy
    Readings: Why, Ch. 1; PCH, Sec. 1.1
    pw: *1zQgX7+
Lecture 2   Structural Causal Models & Causal Diagrams
Jan 8
  • - Structural Causal Models; Causal Diagrams, D-separation.
    Readings: CAI, Ch.2 (Foundations of Causal AI).
    pw: KQX21^R9
Week 2
Lecture 3   1st Layer of PCH
Jan 13
  • - Testable implications of causal diagrams; d-separation.
    pw: EY^!!51K
Lecture 4   2nd Layer of PCH - Identification of Causal Effects
Jan 15
  • - Intuition and Definition of Causal Effects; the Truncated Factorization Product; the Identification Problem
    Readings: CAI, Ch. 4 (Sec. 4.1-4.2).
    pw: pn1=n%l!
Week 3
Lecture 5   Back-Door Criterion
Jan 20
  • - Confounding Bias; Back-Door Criterion; Back-Door Identification
    Readings: CAI, Ch.4 (Sec. 4.1-4.2).
    pw: qV519s%?
Lecture 6   Estimation of Causal Effects - Part I
Jan 22
  • - Inverse Propensity Weighting (IPW); Regression Methods; Augmented IPW; Double Machine Learning
Week 4
Lecture 7   Estimation of Causal Effects - Part II
Jan 27
  • - Estimation of Causal Effects on MIMIC-IV data using IPW, regression, AIPW, and DML.
    pw: Lg09Y=9^
Lecture 8   Beyond the Back-Door – do-Calculus
Jan 29
  • - Front-Door Identification; Three Rules of Do-Calculus.
    Readings: CAI Book, Ch. 4, Sec. 4.3-4.4
    pw: V.vGy0m!
Week 5
Lecture 9   Counterfactuals I
Feb 3
  • - Introduction to Counterfactuals (Layer 3 of Pearl's Causal Hierarchy).
    Readings: CAI Book, Ch. 5, Sec. 5.1
    pw: 8Gycv%h^
Lecture 10   Counterfactuals II
Feb 5
  • - Rules of Operating Over Counterfactuals
    Readings: CAI Book, Ch. 5, Sec. 5.2
    pw: Q81*#*%S
Week 6
Lecture 11   Revision Session
Feb 10
  • - Overview of the concepts studied in the lectures so far.
    pw: r&1Me@?0
Lecture 12   Midterm Exam
Feb 12
  • - Midterm exam.
Week 7
Lecture 13   Causal Effect Heterogeneity
Feb 17
  • - Using counterfactuals for understanding heterogenous effects.
    Readings: CAI Book, Ch. 3, Secs. 3.1-3.2
    pw: GxGiNB#4
Lecture 14   Variation Analysis
Feb 19
  • - Decomposing & Understanding direct, indirect, and spurious effects.
    Readings: CAI Book, Ch. 6
    pw: l.x75ye#
Week 8
Lecture 15   Missing Data - Part I
Feb 24
  • - Using graphical models for understanding missing data.
    pw: ?SbnW%3B
Lecture 16   Missing Data - Part II
Feb 26
  • - Using graphical models for understanding missing data.
    pw: *mw?ie6^
Week 9
Lecture 17   Measurement Error
Mar 3
  • - Using graphical models for measurement error.
    pw: s+.*D8UE
Lecture 18   Unobserved Confounding - Part I
Mar 5
  • - Sensitivity Analysis for Unobserved Confouding.
    pw: THU.E45p
Week 10
Lecture 19   Unobserved Confounding - Part II
Mar 10
  • - Sensitivity Analysis for Unobserved Confouding.
    pw: +4wp%cy2
Lecture 20   Causal Artificial Intelligence
Mar 12
  • - Intersecting Causal Inference with Artificial Intelligence.
    pw: eFwtA*=2

Homeworks

Homework 1 (due Jan 27)
Homework 2 (due Feb 4)
Homework 3 (due Feb 10)
Homework 4 (CITI Certificate) (due Mar 3)

Reading

  • [Why] The Book of Why
    J. Pearl, D. Mackenzie
    Basic Books, 2018.

  • [CAI] Causal Artificial Intelligence  
    E. Bareinboim
    Available online, 2025.

  • [C] Causality: Models, Reasoning, and Inference
    J. Pearl
    Cambridge Press, 2000.

  • [PCH] On Pearl’s Hierarchy and the Foundations of Causal Inference  
    E. Bareinboim, J. Correa, D. Ibeling, T. Icard
    In: “Probabilistic and Causal Inference: The Works of Judea Pearl”, ACM Turing Series, 2020.

Grading

Undergraduate (C160) Graduate (C260)
Homework 10% 10%
Midterm exam (Feb 12) 50% —
Midterm project report (Feb 22) — 30%
Final project (Mar 20) 40% 60%
Graduate students are welcome and encouraged to take the midterm exam; it does not count toward the graduate grade. Please review the UCLA honor code. While working on assignments in small teams is okay, submitted work must be your own.