Causal Discovery in the Quantum Context
Talkat Quantum Physics and Statistical Causal Models (Simons Institute workshop, 2022) · 2022-04-27..04-29·
summary
Classical causal discovery breaks on Bell correlations, so the proposal formalizes quantum causal models via a process matrix W that factorizes into independent quantum channels and a process rule (trace of CP maps ⊗ W) that returns joint probabilities from local interventions. An algorithm using informationally complete instruments recovers causal order and a unique DAG when all quantum systems are included, detects non‑Markovian hidden causes but cannot localize them, and scales exponentially, so ML and tensor‑network heuristics are explored.
AI-written summary · about
chapters (22)
- 0:00Intro
- 1:38Outline
- 2:58Problems
- 5:40Observations vs interventions
- 7:50Quantum engineers
- 14:12Circuit model
- 16:17CP Maps
- 19:15Interventions
- 20:08Mechanisms
- 21:03Causal Structure
- 22:16Process Matrix
- 24:59Process Rule
- 26:53Causal Discovery
- 28:04Causal Discovery Algorithm
- 30:17Quantum Technology
- 33:31Wellcontrolled Experiments
- 36:07Open Questions
- 37:14Goodness of Fit
- 39:10Under Determination
- 41:06Reversibility Statement
- 43:16Limitations
- 44:45Nonmarkovianity
related talks
01h 02mClassical Strategies and Concepts in Causal Discovery
29m🧠🧹 Causality - Mental Hygiene for Data Science
20mHybrid Quantum-Classical Machine Learning with String Diagrams
29mEfficient Discovery of Actual Causality in Stochastic Systems
01h 06mIntroduction to Causal Graphical Models: Graphs, d-separation, do-calculus
01h 01mThe Mathematics of Causal Inference, with Reflections on Machine Learning and the Logic of Science
01h 06mClassical Strategies and Concepts in Causal Discovery
38mCausal Emergence: When Distortions In A Map Obscure The Territory