Removing systematic errors for exoplanet search via latent causes

B. Schölkopf; D. W. Hogg; D. Wang; D. Foreman-Mackey; D. Janzing; C.-J. Simon-Gabriel; Jonas Martin Peters

Removing systematic errors for exoplanet search via latent causes

B. Schölkopf, D. W. Hogg, D. Wang, D. Foreman-Mackey, D. Janzing, C.-J. Simon-Gabriel, Jonas Martin Peters

4 Citations (Scopus)

Abstract

We describe a method for removing the effect of confounders in order to reconstruct a latent quantity of interest. The method, referred to as half-sibling regression, is inspired by recent work in causal inference using additive noise models. We provide a theoretical justification and illustrate the potential of the method in a challenging astronomy application.

Original language	Undefined/Unknown
Title of host publication	Proceedings of the 32nd International Conference on Machine Learning (ICML)
Number of pages	9
Publication date	2015
Pages	2218-2226
Publication status	Published - 2015
Externally published	Yes

Cite this

@inbook{d906eff960e847bb9bc581b8e483e82e,

title = "Removing systematic errors for exoplanet search via latent causes",

abstract = "We describe a method for removing the effect of confounders in order to reconstruct a latent quantity of interest. The method, referred to as half-sibling regression, is inspired by recent work in causal inference using additive noise models. We provide a theoretical justification and illustrate the potential of the method in a challenging astronomy application.",

author = "B. Sch{\"o}lkopf and Hogg, {D. W.} and D. Wang and D. Foreman-Mackey and D. Janzing and C.-J. Simon-Gabriel and Peters, {Jonas Martin}",

year = "2015",

language = "Udefineret/Ukendt",

pages = "2218--2226",

booktitle = "Proceedings of the 32nd International Conference on Machine Learning (ICML)",

}

TY - CHAP

T1 - Removing systematic errors for exoplanet search via latent causes

AU - Schölkopf, B.

AU - Hogg, D. W.

AU - Wang, D.

AU - Foreman-Mackey, D.

AU - Janzing, D.

AU - Simon-Gabriel, C.-J.

AU - Peters, Jonas Martin

PY - 2015

Y1 - 2015

N2 - We describe a method for removing the effect of confounders in order to reconstruct a latent quantity of interest. The method, referred to as half-sibling regression, is inspired by recent work in causal inference using additive noise models. We provide a theoretical justification and illustrate the potential of the method in a challenging astronomy application.

AB - We describe a method for removing the effect of confounders in order to reconstruct a latent quantity of interest. The method, referred to as half-sibling regression, is inspired by recent work in causal inference using additive noise models. We provide a theoretical justification and illustrate the potential of the method in a challenging astronomy application.

M3 - Bidrag til bog/antologi

SP - 2218

EP - 2226

BT - Proceedings of the 32nd International Conference on Machine Learning (ICML)

ER -