WebDec 18, 2024 · The framework of model-X knockoffs provides a flexible tool for exact finite-sample false discovery rate (FDR) control in variable selection. It also completely bypasses the use of conventional p-values, making it especially appealing in high-dimensional nonlinear models. WebWe develop a method for deep learning inference using knockoffs, DeepLINK, to achieve the goal of variable selection with controlled error rate in deep learning models. We show that DeepLINK can also have high power in variable selection with a …
Robust inference with knockoffs (Journal Article) NSF …
WebRobust inference with knockoffs Rina Foygel Barber∗ Emmanuel J. Candès† Richard J. Samworth‡ October 11, 2024 Abstract We consider the variable selection problem, which seeks to identify important variables influencing a response Y out of many candidate features X1,...,X p. We wish to do so while offering finite-sample WebThe knockoff filter is introduced, a new variable selection procedure controlling the FDR in the statistical linear model whenever there are at least as many observations as variables, … barisdale
Robust inference with knockoffs. - Researcher An App For …
WebA knockoff filter for high-dimensional selective inference. RF Barber, EJ Candès. arXiv preprint arXiv:1602.03574, 2016. 152: ... Robust inference with knockoffs. RF Barber, EJ Candès, RJ Samworth. 96: 2024: Global identifiability of linear structural equation models. M Drton, R Foygel, S Sullivant. 93: WebPerformance in terms of false discovery rate, FDR, and power of the multi-environment knockoff filter, MEKF, implemented with joint and separate statistics, compared with the intersection and pooled heuristics for consistent conditional testing on simulated data from many environments, applied to (a) data in which most conditional associations are … suzuki baleno automatic transmission problems