False Discovery Rate Control via Data Splitting
From MaRDI portal
Abstract: Selecting relevant features associated with a given response variable is an important issue in many scientific fields. Quantifying quality and uncertainty of a selection result via false discovery rate (FDR) control has been of recent interest. This paper introduces a way of using data-splitting strategies to asymptotically control the FDR while maintaining a high power. For each feature, the method constructs a test statistic by estimating two independent regression coefficients via data splitting. FDR control is achieved by taking advantage of the statistic's property that, for any null feature, its sampling distribution is symmetric about zero. Furthermore, we propose Multiple Data Splitting (MDS) to stabilize the selection result and boost the power. Interestingly and surprisingly, with the FDR still under control, MDS not only helps overcome the power loss caused by sample splitting, but also results in a lower variance of the false discovery proportion (FDP) compared with all other methods in consideration. We prove that the proposed data-splitting methods can asymptotically control the FDR at any designated level for linear and Gaussian graphical models in both low and high dimensions. Through intensive simulation studies and a real-data application, we show that the proposed methods are robust to the unknown distribution of features, easy to implement and computationally efficient, and are often the most powerful ones amongst competitors especially when the signals are weak and the correlations or partial correlations are high among features.
Cites work
- scientific article; zbMATH DE number 3145583 (Why is no real title available?)
- scientific article; zbMATH DE number 3483405 (Why is no real title available?)
- scientific article; zbMATH DE number 720689 (Why is no real title available?)
- scientific article; zbMATH DE number 1134987 (Why is no real title available?)
- scientific article; zbMATH DE number 1484400 (Why is no real title available?)
- scientific article; zbMATH DE number 3446212 (Why is no real title available?)
- scientific article; zbMATH DE number 845714 (Why is no real title available?)
- A Method to Increase the Power of Multiple Testing Procedures Through Sample Splitting
- A Scale-Free Approach for False Discovery Rate Control in Generalized Linear Models
- A high-dimensional power analysis of the conditional randomization test and knockoffs
- A knockoff filter for high-dimensional selective inference
- A note on data-splitting for the evaluation of significance levels
- A review of Bayesian variable selection methods: what, how and which
- A significance test for the lasso
- A simple forward selection procedure based on false discovery rate control
- Confidence Intervals and Hypothesis Testing for High-Dimensional Regression
- Confidence intervals for low dimensional parameters in high dimensional linear models
- Controlling False Discovery Rate Using Gaussian Mirrors
- Controlling the false discovery rate via knockoffs
- Deep knockoffs
- Derandomizing Knockoffs
- Empirical Bayes Analysis of a Microarray Experiment
- Exact post-selection inference, with application to the Lasso
- False discovery rate control via debiased Lasso
- Gaussian graphical model estimation with false discovery rate control
- Gene hunting with hidden Markov model knockoffs
- Global and Simultaneous Hypothesis Testing for High-Dimensional Logistic Regression Models
- High-dimensional graphs and variable selection with the Lasso
- High-dimensional inference: confidence intervals, \(p\)-values and R-software \texttt{hdi}
- High-dimensional variable selection
- Higher criticism for detecting sparse heterogeneous mixtures.
- Metropolized Knockoff Sampling
- Model Selection and Estimation in Regression with Grouped Variables
- Multilayer knockoff filter: controlled variable selection at multiple resolutions
- On asymptotically optimal confidence regions and tests for high-dimensional models
- On false discovery control under dependence
- On the conditions used to prove oracle results for the Lasso
- Panning for Gold: ‘Model-X’ Knockoffs for High Dimensional Controlled Variable Selection
- RANK: Large-Scale Inference With Graphical Nonlinear Knockoffs
- Relaxing the assumptions of knockoffs by conditioning
- Robust inference with knockoffs
- Robustness of multiple testing procedures against dependence
- SLOPE-adaptive variable selection via convex optimization
- Some results on false discovery rate in stepwise multiple testing procedures.
- Stability selection. With discussion and authors' reply
- Strong Control, Conservative Point Estimation and Simultaneous Conservative Consistency of False Discovery Rates: A Unified Approach
- Testing the prediction error difference between 2 predictors
- The control of the false discovery rate in multiple testing under dependency.
- The positive false discovery rate: A Bayesian interpretation and the \(q\)-value
- Valid post-selection inference
- \(p\)-values for high-dimensional regression
Cited in
(31)- A Scale-Free Approach for False Discovery Rate Control in Generalized Linear Models
- Data-adaptive binary classifiers in high dimensions using random partitioning
- ARK: robust knockoffs inference with coupling
- Scalable Model-Free Feature Screening via Sliced-Wasserstein Dependency
- FDR control and power analysis for high-dimensional logistic regression via Stabkoff
- Comments on “Data Fission: Splitting a Single Data Point” by James Leiner, Boyan Duan, Larry Wasserman, and Aaditya Ramdas
- Discussion of “A Scale-Free Approach for False Discovery Rate Control in Generalized Linear Models”
- Feature screening via false discovery rate control for linear model with multivariate responses
- Ultra-high-dimensional threshold selection for quantile feature screening with false discovery rate error rate control: a case study on high blood pressure analysis
- Model-free group screening and FDR control with deep knockoffs
- Adaptive Selection for False Discovery Rate Control Leveraging Symmetry
- Data Thinning for Poisson Factor Models and its Applications
- False Discovery Rate Smoothing
- Uniform knockoff filter for high-dimensional controlled graph recovery
- Overview of research advance for knockoff methods
- StarTrek: combinatorial variable selection with false discovery rate control
- FDR control for linear log-contrast models with high-dimensional compositional covariates
- Comment on “A Scale-Free Approach for False Discovery Rate Control in Generalized Linear Models” by Chengguang Dai, Buyu Lin, Xin Xing, and Jun S. Liu
- Comments on “A Scale-Free Approach for False Discovery Rate Control in Generalized Linear Models”
- Discussion of “A Scale-Free Approach for False Discovery Rate Control in Generalized Linear Models” by Chenguang Dai, Buyu Lin, Xin Xing, and Jun S. Liu
- Rejoinder: A Scale-free Approach for False Discovery Rate Control in Generalized Linear Models
- A high-dimensional classifier with variable selection using mirror statistics
- Unsupervised curve clustering using wavelets
- Feature screening for clustering analysis of count data with an application to single-cell RNA-sequencing
- A stable and adaptive polygenic signal detection method based on repeated sample splitting
- False discovery rate control for lesion-symptom mapping with heterogeneous data via weighted p-values
- High-dimensional partial linear model with trend filtering
- Comparing LASSO and IPF-LASSO for multi-modal data: variable selection with Type I error control
- False discovery rate control for confounder selection using mirror statistics
- Filtering the Rejection Set While Preserving False Discovery Rate Control
- Calibrated Equilibrium Estimation and Double Selection for High-dimensional Partially Linear Measurement Error Models
This page was built for publication: False Discovery Rate Control via Data Splitting
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6144765)
