Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Entezari, Alireza, Banerjee, Arunava, Kalantari, Leila
Format: Preprint
Veröffentlicht: 2021
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2111.06931
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913868124520448
author Entezari, Alireza
Banerjee, Arunava
Kalantari, Leila
author_facet Entezari, Alireza
Banerjee, Arunava
Kalantari, Leila
contents The method of Alternating Projections (AP) is a fundamental iterative technique with applications to problems in machine learning, optimization and signal processing. Examples include the Gauss-Seidel algorithm which is used to solve large-scale regression problems and the Kaczmarz and projections onto convex sets (POCS) algorithms that are fundamental to iterative reconstruction. Progress has been made with regards to the questions of efficiency and rate of convergence in the randomized setting of the AP method. Here, we extend these results with volume sampling to block (batch) sizes greater than 1 and provide explicit formulas that relate the convergence rate bounds to the spectrum of the underlying system. These results, together with a trace formula and associated volume sampling, prove that convergence rates monotonically improve with larger block sizes, a feature that can not be guaranteed in general with uniform sampling (e.g., in SGD).
format Preprint
id arxiv_https___arxiv_org_abs_2111_06931
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle Kaczmarz and Gauss-Seidel Algorithms with Volume Sampling
Entezari, Alireza
Banerjee, Arunava
Kalantari, Leila
Numerical Analysis
Optimization and Control
52C35, 65F10, 15A18, 52A38
The method of Alternating Projections (AP) is a fundamental iterative technique with applications to problems in machine learning, optimization and signal processing. Examples include the Gauss-Seidel algorithm which is used to solve large-scale regression problems and the Kaczmarz and projections onto convex sets (POCS) algorithms that are fundamental to iterative reconstruction. Progress has been made with regards to the questions of efficiency and rate of convergence in the randomized setting of the AP method. Here, we extend these results with volume sampling to block (batch) sizes greater than 1 and provide explicit formulas that relate the convergence rate bounds to the spectrum of the underlying system. These results, together with a trace formula and associated volume sampling, prove that convergence rates monotonically improve with larger block sizes, a feature that can not be guaranteed in general with uniform sampling (e.g., in SGD).
title Kaczmarz and Gauss-Seidel Algorithms with Volume Sampling
topic Numerical Analysis
Optimization and Control
52C35, 65F10, 15A18, 52A38
url https://arxiv.org/abs/2111.06931