Learning to increase the power of conditional randomization tests

DOI10.1007/S10994-023-06302-3arXiv2207.01022OpenAlexW4318948663MaRDI QIDQ6134320FDOQ6134320

Authors: Shalev Shaer, Yaniv Romano 

Publication date: 22 August 2023

Published in: Machine Learning (Search for Journal in Brave)

Abstract: The model-X conditional randomization test is a generic framework for conditional independence testing, unlocking new possibilities to discover features that are conditionally associated with a response of interest while controlling type-I error rates. An appealing advantage of this test is that it can work with any machine learning model to design powerful test statistics. In turn, the common practice in the model-X literature is to form a test statistic using machine learning models, trained to maximize predictive accuracy with the hope to attain a test with good power. However, the ideal goal here is to drive the model (during training) to maximize the power of the test, not merely the predictive accuracy. In this paper, we bridge this gap by introducing, for the first time, novel model-fitting schemes that are designed to explicitly improve the power of model-X tests. This is done by introducing a new cost function that aims at maximizing the test statistic used to measure violations of conditional independence. Using synthetic and real data sets, we demonstrate that the combination of our proposed loss function with various base predictive models (lasso, elastic net, and deep neural networks) consistently increases the number of correct discoveries obtained, while maintaining type-I error rates under control.

Full work available at URL: https://arxiv.org/abs/2207.01022

Recommendations

• How to gain on power: novel conditional independence tests based on short expansion of conditional mutual information
• Hypothesis testing in adaptively sampled data: ART to maximize power beyond \textit{iid }sampling
• The conditional permutation test for independence while controlling for confounders
• Testing conditional independence via integrating-up transform
• The Hardness of Conditional Independence Testing and the Generalised Covariance Measure

zbMATH Keywords

false discovery rateconditional independence testingconditional randomization testmodel-X knockoffscontrolled feature selection

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05)

Cites Work

• The Hardness of Conditional Independence Testing and the Generalised Covariance Measure
• Causal inference by using invariant prediction: identification and confidence intervals. With discussion and authors' reply
• Distributed optimization and statistical learning via the alternating direction method of multipliers
• Title not available (Why is that?)
• Title not available (Why is that?)
• Regularization and Variable Selection Via the Elastic Net
• The control of the false discovery rate in multiple testing under dependency.
• Variable selection for sparse Dirichlet-multinomial regression with an application to microbiome data analysis
• On the power of conditional independence testing under model-X
• Panning for Gold: ‘Model-X’ Knockoffs for High Dimensional Controlled Variable Selection
• Deep knockoffs
• Gene hunting with hidden Markov model knockoffs
• Controlling the false discovery rate via knockoffs
• Title not available (Why is that?)
• The Holdout Randomization Test for Feature Selection in Black Box Models
• Causal inference in genetic trio studies
• Fast and powerful conditional randomization testing via distillation

Cited In (1)

• Hypothesis testing in adaptively sampled data: ART to maximize power beyond \textit{iid }sampling

This page was built for publication: Learning to increase the power of conditional randomization tests

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6134320)