Three-quarter Sibling Regression for Denoising Observational Data
This method addresses the problem of denoising observational data for ecological studies and conservation policies, particularly when dealing with dependent variables controlled by common causes.
This paper introduces 'three-quarter sibling regression' to filter systematic noise in observational data when latent variables have observed common causes. It was demonstrated to reduce systematic detection variability caused by moon brightness in moth surveys.
Many ecological studies and conservation policies are based on field observations of species, which can be affected by systematic variability introduced by the observation process. A recently introduced causal modeling technique called 'half-sibling regression' can detect and correct for systematic errors in measurements of multiple independent random variables. However, it will remove intrinsic variability if the variables are dependent, and therefore does not apply to many situations, including modeling of species counts that are controlled by common causes. We present a technique called 'three-quarter sibling regression' to partially overcome this limitation. It can filter the effect of systematic noise when the latent variables have observed common causes. We provide theoretical justification of this approach, demonstrate its effectiveness on synthetic data, and show that it reduces systematic detection variability due to moon brightness in moth surveys.