Logistic regression with total variation regularization

zbMATH Open1458.62153arXiv2003.02678MaRDI QIDQ5146339FDOQ5146339

Authors: Sara Van De Geer 

Publication date: 25 January 2021

Abstract: We study logistic regression with total variation penalty on the canonical parameter and show that the resulting estimator satisfies a sharp oracle inequality: the excess risk of the estimator is adaptive to the number of jumps of the underlying signal or an approximation thereof. In particular when there are finitely many jumps, and jumps up are sufficiently separated from jumps down, then the estimator converges with a parametric rate up to a logarithmic term $logn/n$, provided the tuning parameter is chosen appropriately of order $1/sqrtn$. Our results extend earlier results for quadratic loss to logistic loss. We do not assume any a priori known bounds on the canonical parameter but instead only make use of the local curvature of the theoretical risk.

Full work available at URL: https://arxiv.org/abs/2003.02678

Recommendations

• On the total variation regularized estimator over a class of tree graphs
• Logistic Discrimination with Total Variation Regularization
• scientific article; zbMATH DE number 6179253
• Self-concordant analysis for logistic regression
• Hard thresholding regularised logistic regression: theory and algorithms

zbMATH Keywords

logistic regressionoracle inequalitytotal variation penalty

Mathematics Subject Classification ID

Nonparametric estimation (62G05) Ridge regression; shrinkage estimators (Lasso) (62J07) Generalized linear models (logistic models) (62J12)

Cites Work

• Nonlinear total variation based noise removal algorithms
• Weak convergence and empirical processes. With applications to statistics
• Adaptive piecewise polynomial estimation via trend filtering
• The DFS fused Lasso: linear-time denoising over general graphs
• Sparsity and Smoothness Via the Fused Lasso
• Adaptive risk bounds in univariate total variation denoising and trend filtering
• Estimation and testing under sparsity. École d'Été de Probabilités de Saint-Flour XLV -- 2015
• Additive models with trend filtering
• On the prediction performance of the Lasso
• Splines in higher order TV regularization
• On the total variation regularized estimator over a class of tree graphs

Cited In (5)

• Logistic classification with varying gaussians
• Logistic regression, AdaBoost and Bregman distances
• Logistic Discrimination with Total Variation Regularization
• Prediction bounds for higher order total variation regularized least squares
• Self-concordant analysis for logistic regression

This page was built for publication: Logistic regression with total variation regularization

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5146339)