Learning with optimal interpolation norms

DOI10.1007/S11075-018-0568-1zbMATH Open1454.90049arXiv1603.09273OpenAlexW2753852770WikidataQ129562610 ScholiaQ129562610MaRDI QIDQ2420165FDOQ2420165

Authors: Patrick L. Combettes, Andrew M. Mcdonald, Charles A. Micchelli, Massimiliano Pontil 

Publication date: 5 June 2019

Published in: Numerical Algorithms (Search for Journal in Brave)

Abstract: We analyze a class of norms defined via an optimal interpolation problem involving the composition of norms and a linear operator. This construction, known as infimal postcomposition in convex analysis, is shown to encompass various of norms which have been used as regularizers in machine learning, signal processing, and statistics. In particular, these include the latent group lasso, the overlapping group lasso, and certain norms used for learning tensors. We establish basic properties of this class of norms and we provide dual norms. The extension to more general classes of convex functions is also discussed. A stochastic block-coordinate version of the Douglas-Rachford algorithm is devised to solve minimization problems involving these regularizers. A prominent feature of the algorithm is that it yields iterates that converge to a solution in the case of non smooth losses and random block updates. Finally, we present numerical experiments with problems employing the latent group lasso penalty.

Full work available at URL: https://arxiv.org/abs/1603.09273

Recommendations

• scientific article; zbMATH DE number 6253954
• Structured sparsity via alternating direction methods
• Proximal methods for the latent group lasso penalty
• Structured variable selection with sparsity-inducing norms
• Structured sparsity through convex optimization

zbMATH Keywords

machine learningDouglas-Rachford splittingblock-coordinate proximal algorithminfimal postcompositionlatent group lassooptimal interpolation norm

Mathematics Subject Classification ID

Convex programming (90C25) Monotone operators and generalizations (47H05) Programming in abstract spaces (90C48)

Cites Work

• Sparsity and Smoothness Via the Fused Lasso
• Model Selection and Estimation in Regression with Grouped Variables
• Proximal methods for hierarchical sparse coding
• Tensor Decompositions and Applications
• The composite absolute penalties family for grouped and hierarchical variable selection
• Convex Analysis
• Title not available (Why is that?)
• Optimization with sparsity-inducing penalties
• Proximity algorithms for image models: denoising
• Tensor completion and low-\(n\)-rank tensor recovery via convex optimization
• Title not available (Why is that?)
• A new approach in interpolation spaces
• Learning the kernel function via regularization
• Systems of Structured Monotone Inclusions: Duality, Algorithms, and Applications
• Stochastic quasi-Fejér block-coordinate fixed point iterations with random sweeping
• The convex geometry of linear inverse problems
• Using Block Norms for Location Modeling
• Convex multi-task feature learning
• Learning with tensors: a framework based on convex optimization and spectral regularization
• Convex analysis and monotone operator theory in Hilbert spaces
• An algorithm for splitting parallel sums of linearly composed monotone operators, with applications to signal recovery
• Super-Resolution With Sparse Mixing Estimators
• Title not available (Why is that?)
• Feature space perspectives for learning the kernel
• Robust 2D Principal Component Analysis: A Structured Sparsity Regularized Approach
• Fundamental Performance Limits for Ideal Decoders in High-Dimensional Linear Inverse Problems
• Background Subtraction Based on Low-Rank and Structured Sparse Decomposition
• On spectral learning
• Prolongements de foncteurs d'interpolation et applications
• New perspectives on \(k\)-support and cluster norms
• Exploring Structured Sparsity by a Reweighted Laplace Prior for Hyperspectral Compressive Sensing
• Stochastic quasi-Fejér block-coordinate fixed point iterations with random sweeping. II: Mean-square and linear convergence
• Regularizers for structured sparsity
• Structured sparsity and generalization

Cited In (1)

• Learning the Dynamics of Objects by Optimal Functional Interpolation

This page was built for publication: Learning with optimal interpolation norms

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2420165)