Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Mo, Kai-Chia, Shalev-Shwartz, Shai, Shártov, Nisæl
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2405.04710
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910907722891264
author Mo, Kai-Chia
Shalev-Shwartz, Shai
Shártov, Nisæl
author_facet Mo, Kai-Chia
Shalev-Shwartz, Shai
Shártov, Nisæl
contents We describe an apparatus for subgradient-following of the optimum of convex problems with variational penalties. In this setting, we receive a sequence $y_i,\ldots,y_n$ and seek a smooth sequence $x_1,\ldots,x_n$. The smooth sequence needs to attain the minimum Bregman divergence to an input sequence with additive variational penalties in the general form of $\sum_i{}g_i(x_{i+1}-x_i)$. We derive known algorithms such as the fused lasso and isotonic regression as special cases of our approach. Our approach also facilitates new variational penalties such as non-smooth barrier functions. We then derive a novel lattice-based procedure for subgradient following of variational penalties characterized through the output of arbitrary convolutional filters. This paradigm yields efficient solvers for high-order filtering problems of temporal sequences in which sparse discrete derivatives such as acceleration and jerk are desirable. We also introduce and analyze new multivariate problems in which $\mathbf{x}_i,\mathbf{y}_i\in\mathbb{R}^d$ with variational penalties that depend on $\|\mathbf{x}_{i+1}-\mathbf{x}_i\|$. The norms we consider are $\ell_2$ and $\ell_\infty$ which promote group sparsity.
format Preprint
id arxiv_https___arxiv_org_abs_2405_04710
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Untangling Lariats: Subgradient Following of Variationally Penalized Objectives
Mo, Kai-Chia
Shalev-Shwartz, Shai
Shártov, Nisæl
Machine Learning
Optimization and Control
We describe an apparatus for subgradient-following of the optimum of convex problems with variational penalties. In this setting, we receive a sequence $y_i,\ldots,y_n$ and seek a smooth sequence $x_1,\ldots,x_n$. The smooth sequence needs to attain the minimum Bregman divergence to an input sequence with additive variational penalties in the general form of $\sum_i{}g_i(x_{i+1}-x_i)$. We derive known algorithms such as the fused lasso and isotonic regression as special cases of our approach. Our approach also facilitates new variational penalties such as non-smooth barrier functions. We then derive a novel lattice-based procedure for subgradient following of variational penalties characterized through the output of arbitrary convolutional filters. This paradigm yields efficient solvers for high-order filtering problems of temporal sequences in which sparse discrete derivatives such as acceleration and jerk are desirable. We also introduce and analyze new multivariate problems in which $\mathbf{x}_i,\mathbf{y}_i\in\mathbb{R}^d$ with variational penalties that depend on $\|\mathbf{x}_{i+1}-\mathbf{x}_i\|$. The norms we consider are $\ell_2$ and $\ell_\infty$ which promote group sparsity.
title Untangling Lariats: Subgradient Following of Variationally Penalized Objectives
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2405.04710