Comparer des méthodes

	Algorithmes de découverte causale (PC, FCI, LiNGAM)	Identification causale avec les graphes acycliques dirigés (do-calculus)	Variables instrumentales par moindres carrés en deux étapes (VI/2SLS)
Domaine	Inférence causale	Inférence causale	Inférence causale
Famille	Regression model	Regression model	Regression model
Année d'origine	2000	2009	2009
Auteur d'origine	Spirtes, Glymour & Scheines (PC/FCI); Shimizu et al. (LiNGAM)	Judea Pearl	Angrist & Pischke (textbook treatment); Stock & Yogo (weak-instrument theory)
Type	Causal structure learning	Causal identification framework	Instrumental-variables regression
Source fondatrice	Spirtes, P., Glymour, C., & Scheines, R. (2000). Causation, Prediction, and Search (2nd ed.). MIT Press. ISBN: 978-0262194402	Pearl, J. (2009). Causality: Models, Reasoning, and Inference (2nd ed.). Cambridge University Press. ISBN: 978-0521895606	Angrist, J. D. & Pischke, J. S. (2009). Mostly Harmless Econometrics: An Empiricist's Companion. Princeton University Press. ISBN: 978-0691120355
Alias	PC algorithm, FCI algorithm, LiNGAM, causal structure learning	do-calculus, backdoor adjustment, Pearl causal identification, DAG ile Nedensel Tanımlama (do-calculus)	instrumental variables, IV estimation, 2SLS, instrumental variable regression
Apparentées	5	5	5
Résumé	Causal discovery is a family of algorithms that automatically learn a directed acyclic graph (DAG) describing causal structure directly from observational data. The constraint-based PC and FCI algorithms were developed by Spirtes, Glymour and Scheines (2000), while the LiNGAM model of Shimizu et al. (2006) exploits linear non-Gaussian structure to orient edges.	DAG causal identification is a framework, developed by Judea Pearl (2009), that encodes causal assumptions as a directed acyclic graph and uses the do-calculus rules to determine whether and how a causal effect can be identified from observational data. It systematically handles confounders, instrumental variables, and backdoor paths.	IV/2SLS is a two-stage estimation method that recovers the causal effect of an endogenous regressor by isolating the part of its variation driven by an external instrument. It is the workhorse identification strategy in modern applied econometrics, developed at length in Angrist and Pischke's Mostly Harmless Econometrics (2009).