Detecting low-complexity unobserved causes

D. Janzing; E. Sgouritsa; O. Stegle; Jonas Martin Peters; B. Schölkopf

Detecting low-complexity unobserved causes

D. Janzing, E. Sgouritsa, O. Stegle, Jonas Martin Peters, B. Schölkopf

8 Citations (Scopus)

Abstract

We describe a method that infers whether statistical dependences between two observed variables X and Y are due to a \direct" causal link or only due to a connecting causal path that contains an unobserved variable of low complexity, e.g., a binary variable. This problem is motivated by statistical genetics. Given a genetic marker that is correlated with a phenotype of interest, we want to detect whether this marker is causal or it only correlates with a causal one. Our method is based on the analysis of the location of the conditional distributions P(Y /x) in the simplex of all distributions of Y . We report encouraging results on semi-empirical data.

Original language	Undefined/Unknown
Title of host publication	Proceedings of the 27th Annual Conference on Uncertainty in Artificial Intelligence (UAI)
Number of pages	9
Publication date	2011
Pages	383-391
Publication status	Published - 2011
Externally published	Yes

Cite this

@inbook{2f10534592604c51880065759bb28ba9,

title = "Detecting low-complexity unobserved causes",

abstract = "We describe a method that infers whether statistical dependences between two observed variables X and Y are due to a \direct{"} causal link or only due to a connecting causal path that contains an unobserved variable of low complexity, e.g., a binary variable. This problem is motivated by statistical genetics. Given a genetic marker that is correlated with a phenotype of interest, we want to detect whether this marker is causal or it only correlates with a causal one. Our method is based on the analysis of the location of the conditional distributions P(Y /x) in the simplex of all distributions of Y . We report encouraging results on semi-empirical data.",

author = "D. Janzing and E. Sgouritsa and O. Stegle and Peters, {Jonas Martin} and B. Sch{\"o}lkopf",

year = "2011",

language = "Udefineret/Ukendt",

pages = "383--391",

booktitle = "Proceedings of the 27th Annual Conference on Uncertainty in Artificial Intelligence (UAI)",

}

TY - CHAP

T1 - Detecting low-complexity unobserved causes

AU - Janzing, D.

AU - Sgouritsa, E.

AU - Stegle, O.

AU - Peters, Jonas Martin

AU - Schölkopf, B.

PY - 2011

Y1 - 2011

N2 - We describe a method that infers whether statistical dependences between two observed variables X and Y are due to a \direct" causal link or only due to a connecting causal path that contains an unobserved variable of low complexity, e.g., a binary variable. This problem is motivated by statistical genetics. Given a genetic marker that is correlated with a phenotype of interest, we want to detect whether this marker is causal or it only correlates with a causal one. Our method is based on the analysis of the location of the conditional distributions P(Y /x) in the simplex of all distributions of Y . We report encouraging results on semi-empirical data.

AB - We describe a method that infers whether statistical dependences between two observed variables X and Y are due to a \direct" causal link or only due to a connecting causal path that contains an unobserved variable of low complexity, e.g., a binary variable. This problem is motivated by statistical genetics. Given a genetic marker that is correlated with a phenotype of interest, we want to detect whether this marker is causal or it only correlates with a causal one. Our method is based on the analysis of the location of the conditional distributions P(Y /x) in the simplex of all distributions of Y . We report encouraging results on semi-empirical data.

M3 - Bidrag til bog/antologi

SP - 383

EP - 391

BT - Proceedings of the 27th Annual Conference on Uncertainty in Artificial Intelligence (UAI)

ER -