Adaptive variable selection in nonparametric sparse additive models

DOI10.1214/17-EJS1275MaRDI QIDQ2396347zbMATH OpenOpenAlexFDO

Authors Cristina Butucea, Natalia Stepanova

Publication date 8 June 2017

Published in Electronic Journal of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1508.06660, https://projecteuclid.org/euclid.ejs/1495850627

zbMATH Keywords

adaptive variable selectionexact and almost full selectorshigh-dimensional nonparametric regressionsparse additive signals

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Asymptotic properties of nonparametric inference (62G20)

Abstract: We consider the problem of recovery of an unknown multivariate signal $f$ observed in a $d$-dimensional Gaussian white noise model of intensity $varepsilon$. We assume that $f$ belongs to a class of smooth functions ${calF}^{d}subset{L}_2([0,1{]}^d)$ and has an additive sparse structure determined by the parameter $s$, the number of non-zero univariate components contributing to $f$. We are interested in the case when $d=d_{v}arepsilonoinfty$ as $varepsilono0$ and the parameter $s$ stays "small" relative to $d$. With these assumptions, the recovery problem in hand becomes that of determining which sparse additive components are non-zero. Attempting to reconstruct most non-zero components of $f$, but not all of them, we arrive at the problem of almost full variable selection in high-dimensional regression. For two different choices of ${calF}^d$, we establish conditions under which almost full variable selection is possible, and provide a procedure that gives almost full variable selection. The procedure does the best (in the asymptotically minimax sense) in selecting most non-zero components of $f$. Moreover, it is adaptive in the parameter $s$.

Recommendations

• Adaptive variable selection in nonparametric sparse regression
• Flexible variable selection for recovering sparsity in nonadditive nonparametric models
• Variable selection in nonparametric additive models
• Nonparametric variable selection and its application to additive models
• Variable selection for partially linear models via adaptive LASSO
• Variable selection using adaptive nonlinear interaction structures in high dimensions
• Adaptive sparse regression
• An Adaptive Method of Variable Selection in Regression
• The adaptive Lasso in high-dimensional sparse heteroscedastic models

Cited in

(15)

• Variable selection with Hamming loss
• Detection of sparse additive functions
• Variable selection consistency of Gaussian process regression
• Uncertainty quantification for robust variable selection and multiple testing
• Adaptive threshold-based classification of sparse high-dimensional data
• Adaptive variable selection in nonparametric sparse regression
• Randomized pick-freeze for sparse Sobol indices estimation in high dimension
• Flexible variable selection for recovering sparsity in nonadditive nonparametric models
• Variable selection, monotone likelihood ratio and group sparsity
• Adaptive signal recovery with Subbotin noise
• Sharp multiple testing boundary for sparse sequences
• Exact recovery in the double sparse model: sufficient and necessary signal conditions
• Adaptive exact recovery in sparse nonparametric models
• Exact variable selection in sparse nonparametric models
• Detection of a sparse variable function

This page was built for publication: Adaptive variable selection in nonparametric sparse additive models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2396347)