Deep latent variable models for generating knockoffs

Ying, Liu; Zheng, Cheng

doi:10.1002/sta4.260

Cited by 9 publications

(8 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Romano, Sesia, and Candès (2020) developed a Deep Knockoff machine using deep generative models. Liu and Zheng (2019) developed a Model-X generating method using Deep latent variable models. More recently, Bates, Candès, Janson, and Wang (2020) proposed an efficient general metropolized knockoff sampler.…”

Section: Knockoff-based Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Multiple Testing for Composite Null with FDR Control Guarantee

Dai¹,

Cheng²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

False discovery rate (FDR) controlling procedures provide important statistical guarantees for reproducibility in signal identification experiments with multiple hypotheses testing. In many recent applications, the same set of candidate features are studied in multiple independent experiments. For example, experiments repeated at different facilities and with different cohorts, and association studies with the same candidate features but different outcomes of interest. These studies provide us opportunities to identify signals by considering the experiments jointly. We study the question of how to provide reproducibility guarantees when we test composite null hypotheses on multiple features. Specifically, we test the unions of the null hypotheses from multiple experiments. We present a knockoff-based variable selection method to identify mutual signals from multiple independent experiments, with a finite sample size FDR control guarantee. We demonstrate the performance of this method with numerical studies and applications in analyzing crime data and TCGA data.

show abstract

Section: Knockoff-based Methodsmentioning

confidence: 99%

“…where Swap(j) stands for exchanging the j-th column and the (j +p)-th column. X j can be sampled using various algorithms (Bates et al, 2020;Liu & Zheng, 2019;Romano et al, 2020;Spector & Janson, 2020).…”

Section: Notationsmentioning

confidence: 99%

Multiple Testing for Composite Null with FDR Control Guarantee

Dai¹,

Cheng²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…(5) where Swap(j) stands for exchanging the 𝑗th column and the (𝑗 + 𝑝)th column of [𝐗 X], and 𝐴 d = 𝐵 indicates 𝐴 and 𝐵 are identical in distribution. X can be generated using various algorithms (Barber & Candés, 2015;Romano et al, 2020;Liu & Zheng, 2019;Bates et al, 2021;Spector & Janson, 2022). More knockoff construction details are reviewed in Web Appendix A.1.…”

Section: Algorithmmentioning

confidence: 99%

False Discovery Rate-Controlled Multiple Testing for Union Null Hypotheses: A Knockoff-Based Approach

Dai

Zheng

2023

Biometrics

Self Cite

View full text Add to dashboard Cite

False discovery rate (FDR) controlling procedures provide important statistical guarantees for replicability in signal identification based on multiple hypotheses testing. In many fields of study, FDR controling procedures are used in high‐dimensional (HD) analyses to discover features that are truly associated with the outcome. In some recent applications, data on the same set of candidate features are independently collected in multiple different studies. For example, gene expression data are collected at different facilities and with different cohorts, to identify the genetic biomarkers of multiple types of cancers. These studies provide us with opportunities to identify signals by considering information from different sources (with potential heterogeneity) jointly. This paper is about how to provide FDR control guarantees for the tests of union null hypotheses of conditional independence. We present a knockoff‐based variable selection method (Simultaneous knockoffs) to identify mutual signals from multiple independent datasets, providing exact FDR control guarantees under finite sample settings. This method can work with very general model settings and test statistics. We demonstrate the performance of this method with extensive numerical studies and two real‐data examples.

show abstract

“…Similarly, Bayesian approaches to feature selection [28] have also been generalised to the group setting [35]. Finally, the Knockoff procedure [9,17,39,57,73,80] is a generative procedure that creates fake covariates (knockoffs), obeying certain symmetries under permutations of real and knockoff features. By subsequently carrying out feature selection on the combined real and knockoff data, it is possible to obtain guarantees on the False Discovery Rate of the selected features.…”

Section: Related Workmentioning

confidence: 99%

Composite Feature Selection using Deep Ensembles

Imrie¹,

Alexander²,

Lió³

et al. 2022

Preprint

View full text Add to dashboard Cite

In many real world problems, features do not act alone but in combination with each other. For example, in genomics, diseases might not be caused by any single mutation but require the presence of multiple mutations. Prior work on feature selection either seeks to identify individual features or can only determine relevant groups from a predefined set. We investigate the problem of discovering groups of predictive features without predefined grouping. To do so, we define predictive groups in terms of linear and non-linear interactions between features. We introduce a novel deep learning architecture that uses an ensemble of feature selection models to find predictive groups, without requiring candidate groups to be provided. The selected groups are sparse and exhibit minimum overlap. Furthermore, we propose a new metric to measure similarity between discovered groups and the ground truth. We demonstrate the utility of our model on multiple synthetic tasks and semi-synthetic chemistry datasets, where the ground truth structure is known, as well as an image dataset and a real-world cancer dataset.

show abstract

Deep latent variable models for generating knockoffs

Cited by 9 publications

References 12 publications

Multiple Testing for Composite Null with FDR Control Guarantee

Multiple Testing for Composite Null with FDR Control Guarantee

False Discovery Rate-Controlled Multiple Testing for Union Null Hypotheses: A Knockoff-Based Approach

Composite Feature Selection using Deep Ensembles

Contact Info

Product

Resources

About