An Introduction to Potential Outcomes, DAGs and Single-World Intervention Graphs
summary
Presents a complete, constructive identification procedure: re-express P(Y|do(X)) as a sum of district-specific intervention kernels, then test each kernel for identifiability by iteratively applying a graph operation called the fixing operation that cuts incoming arrows and reweights kernels. Delivers a compact one-line ID algorithm (ancestors-of-Y after removing X-edges, check that all districts are intrinsic sets), shows worked front-door and sequential-randomization examples, and gives a simple hedge-based criterion proving nonidentifiability when the test fails.
AI-written summary · about
chapters (15)
- 0:00Start
- 2:54General Identification Problem
- 3:13Intervention Distributions
- 3:26Intervention Distribution
- 11:43Identification Question
- 13:41Latent Projection
- 17:46Lane Projection
- 20:38Notes of Ancestral Descendants
- 24:15Fixed and Random Nodes
- 25:07Schmitzer Pearl Algorithm
- 27:12Decompose the Query
- 35:05Fixable Vertices
- 38:11Examples
- 51:10One Line Id Algorithm
- 55:13Nested Markov Model
related talks
28mLearning Causal Representations From Unknown Interventions
31mIntroduction to Causal Inference using pgmpy
49mCausal Inference in Complex Systems: Network Interference, Strategic Agents, and Beyond
24mA Friendly Introduction To Causality
44mCausal Reductions
41mOn the Path to Causal Inference
31mCollaborative Causal Discovery With Atomic Interventions
31mPanel Discussion
