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.

Extended version of a paper appearing in the Proceedings of the 32nd International Conference on Machine Learning, Lille, France, 2015