Density-sensitive semisupervised inference

DOI10.1214/13-AOS1092zbMATH Open1267.62057MaRDI QIDQ355103FDOQ355103

Authors Martin Azizyan, Aarti Singh, Larry Wasserman

Publication date 24 July 2013

Published in The Annals of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1204.1685

zbMATH Keywords

efficiencykernel densitynonparametric inferencesemisupervised

Mathematics Subject Classification ID

Density estimation (62G07) Nonparametric regression and quantile regression (62G08) Nonparametric tolerance and confidence regions (62G15)

Abstract: Semisupervised methods are techniques for using labeled data $(X_1,Y_1),ldots,(X_n,Y_n)$ together with unlabeled data $X_{n+1},ldots,X_N$ to make predictions. These methods invoke some assumptions that link the marginal distribution $P_X$ of X to the regression function f(x). For example, it is common to assume that f is very smooth over high density regions of $P_X$. Many of the methods are ad-hoc and have been shown to work in specific examples but are lacking a theoretical foundation. We provide a minimax framework for analyzing semisupervised methods. In particular, we study methods based on metrics that are sensitive to the distribution $P_X$. Our model includes a parameter $alpha$ that controls the strength of the semisupervised assumption. We then use the data to adapt to $alpha$.

Recommendations

• Semi-supervised learning based on high density region estimation
• On semi-supervised learning
• Semi-supervised learning with density-ratio estimation
• A unified view of density-based methods for semi-supervised clustering and classification
• Generalization error bounds in semi-supervised classification under the cluster assumption

Cites work

• A distribution-free theory of nonparametric regression
• Density-sensitive semisupervised inference
• Generalization error bounds in semi-supervised classification under the cluster assumption
• Manifold regularization and semi-supervised learning: some theoretical analyses
• On propagated scoring for semisupervised additive models
• On the Tchebychef Inequality of Bernstein
• Semi-supervised learning on Riemannian manifolds
• Signal Reconstruction From Noisy Random Projections
• Spectral connectivity analysis
• The relative value of labeled and unlabeled samples in pattern recognition with an unknown mixing parameter
• The use of unlabeled data in predictive modeling

Cited in

(12)

• Metric-based methods for adaptive model selection and regularization
• On semi-supervised learning
• Balancing geometry and density: path distances on high-dimensional data
• Conditional probability estimation based classification with class label missing at random
• Minimax semi-supervised set-valued approach to multi-class classification
• Semi-supervised learning with density-ratio estimation
• On propagated scoring for semisupervised additive models
• Semi-supervised inference: general theory and estimation of means
• Density-sensitive semisupervised inference
• Semi-supervised learning based on high density region estimation
• On regression and classification with possibly missing response variables in the data
• A unified view of density-based methods for semi-supervised clustering and classification

Describes a project that uses

Uses Software

• spa

This page was built for publication: Density-sensitive semisupervised inference

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q355103)