Sparse single-index model

MaRDI QIDQ5405247zbMATH OpenFDO

Authors Pierre Alquier, Gérard Biau

Publication date 1 April 2014

Full work available at URL https://arxiv.org/abs/1101.3229, http://www.jmlr.org/papers/v14/alquier13a.html

zbMATH Keywords

oracle inequalityPAC-Bayesianregression estimationreversible jump Markov chain Monte Carlo methodsingle-index modelsparsity

Mathematics Subject Classification ID

Bayesian inference (62F15) Nonparametric regression and quantile regression (62G08) Ridge regression; shrinkage estimators (Lasso) (62J07) Generalized linear models (logistic models) (62J12)

Abstract: Let $$ be a random pair taking values in $mathbb{R}^{p}imesmathbbR$. In the so-called single-index model, one has $$, where $f^{star}$ is an unknown univariate measurable function, $het{a}^{star}$ is an unknown vector in $mathbb{R}^d$, and $W$ denotes a random noise satisfying $$. The single-index model is known to offer a flexible way to model a variety of high-dimensional real-world phenomena. However, despite its relative simplicity, this dimension reduction scheme is faced with severe complications as soon as the underlying dimension becomes larger than the number of observations ("$p$ larger than $n$" paradigm). To circumvent this difficulty, we consider the single-index model estimation problem from a sparsity perspective using a PAC-Bayesian approach. On the theoretical side, we offer a sharp oracle inequality, which is more powerful than the best known oracle inequalities for other common procedures of single-index recovery. The proposed method is implemented by means of the reversible jump Markov chain Monte Carlo technique and its performance is compared with that of standard procedures.

Recommendations

• scientific article; zbMATH DE number 2148835
• Bayesian estimation and variable selection for single index models
• Signed support recovery for single index models in high-dimensions
• Nonconcave penalized inverse regression in single-index models with high dimensional predic\-tors
• A dimension reduction based approach for estimation and variable selection in partially linear single-index models with high-dimensional covariates

Cited in

(37)

• PAC-Bayesian high dimensional bipartite ranking
• Optimal estimation of slope vector in high-dimensional linear transformation models
• A quasi-Bayesian perspective to online clustering
• On the exponentially weighted aggregate with the Laplace prior
• PAC-Bayesian estimation and prediction in sparse additive models
• High dimensional single-index Bayesian modeling of brain atrophy
• Convergence guarantee for the sparse monotone single index model
• A Bayesian approach for noisy matrix completion: optimal rate under general sampling distribution
• Misspecified nonconvex statistical optimization for sparse phase retrieval
• Minimax adaptive dimension reduction for regression
• On some recent advances on high dimensional Bayesian statistics
• On the properties of variational approximations of Gibbs posteriors
• Nonlinear generalization of the monotone single index model
• High-dimensional estimation with geometric constraints
• Scalable approximations for generalized linear problems
• Metric Learning via Cross-Validation
• Ultra-high dimensional single-index quantile regression
• scientific article; zbMATH DE number 7306923 (Why is no real title available?)
• On oracle property and asymptotic validity of Bayesian generalized method of moments
• Structured Ultrahigh Dimensional Multiple-Index Models with Efficient Estimation in Computation And Theory
• User-friendly Introduction to PAC-Bayes Bounds
• Estimation in monotone single‐index models
• High-dimensional sparse single-index regression via Hilbert-Schmidt independence criterion
• Isotonic regression meets Lasso
• Signed support recovery for single index models in high-dimensions
• A sparse enhanced indexation model with chance and cardinality constraints
• Distributed semi-supervised single-index model with corruption
• Multivariate Gaussian RBF-net for smooth function estimation and variable selection
• High-dimensional prediction for count response via sparse exponential weights
• Spectral estimators for structured generalized linear models via approximate message passing
• Broadcasted nonparametric tensor regression
• Solving the missing at random problem in semi-supervised learning: an inverse probability weighting method
• Rates of convergence of the constrained least squares estimator in high-dimensional monotone single-index models
• Generalization bounds: perspectives from information theory and PAC-Bayes
• Kullback-Leibler excess risk bounds for exponential weighted aggregation in generalized linear models
• PAC-Bayes estimation for high-dimensional multi-index models with unknown active dimension
• Isotonic single-index model for high-dimensional database marketing

This page was built for publication: Sparse single-index model

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5405247)