Sibling Regression for Generalized Linear Models

07/03/2021
by   Shiv Shankar, et al.
7

Field observations form the basis of many scientific studies, especially in ecological and social sciences. Despite efforts to conduct such surveys in a standardized way, observations can be prone to systematic measurement errors. The removal of systematic variability introduced by the observation process, if possible, can greatly increase the value of this data. Existing non-parametric techniques for correcting such errors assume linear additive noise models. This leads to biased estimates when applied to generalized linear models (GLM). We present an approach based on residual functions to address this limitation. We then demonstrate its effectiveness on synthetic data and show it reduces systematic detection variability in moth surveys.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
12/31/2020

Three-quarter Sibling Regression for Denoising Observational Data

Many ecological studies and conservation policies are based on field obs...
research
04/09/2021

Measurement Errors in Semiparametric Generalized Linear Models

Regression models that ignore measurement error in predictors may produc...
research
05/13/2019

Hierarchical approaches for flexible and interpretable binary regression models

Binary regression models are ubiquitous in virtually every scientific fi...
research
07/08/2020

Non-parametric Models for Non-negative Functions

Linear models have shown great effectiveness and flexibility in many fie...
research
12/26/2019

A Simple Correction Procedure for High-Dimensional Generalized Linear Models with Measurement Error

We consider high-dimensional generalized linear models when the covariat...
research
12/26/2019

A Simple Correction Procedure for High-Dimensional General Linear Models with Measurement Error

We consider high-dimensional generalized linear models when the covariat...
research
10/13/2016

Removal of Batch Effects using Distribution-Matching Residual Networks

Sources of variability in experimentally derived data include measuremen...

Please sign up or login with your details

Forgot password? Click here to reset