Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Huang, Xiaoqing, Ang, Andersen, Huang, Kun, Zhang, Jie, Wang, Yijie
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2304.05223
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929674397941760
author Huang, Xiaoqing
Ang, Andersen
Huang, Kun
Zhang, Jie
Wang, Yijie
author_facet Huang, Xiaoqing
Ang, Andersen
Huang, Kun
Zhang, Jie
Wang, Yijie
contents We study estimation of piecewise smooth signals over a graph. We propose a $\ell_{2,0}$-norm penalized Graph Trend Filtering (GTF) model to estimate piecewise smooth graph signals that exhibit inhomogeneous levels of smoothness across the nodes. We prove that the proposed GTF model is simultaneously a k-means clustering on the signal over the nodes and a minimum graph cut on the edges of the graph, where the clustering and the cut share the same assignment matrix. We propose two methods to solve the proposed GTF model: a spectral decomposition method and a method based on simulated annealing. In the experiment on synthetic and real-world datasets, we show that the proposed GTF model has a better performances compared with existing approaches on the tasks of denoising, support recovery and semi-supervised classification. We also show that the proposed GTF model can be solved more efficiently than existing models for the dataset with a large edge set.
format Preprint
id arxiv_https___arxiv_org_abs_2304_05223
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Inhomogeneous graph trend filtering via a l2,0 cardinality penalty
Huang, Xiaoqing
Ang, Andersen
Huang, Kun
Zhang, Jie
Wang, Yijie
Machine Learning
Social and Information Networks
65F50, 68U01, 68R01
G.1.6; G.1.10
We study estimation of piecewise smooth signals over a graph. We propose a $\ell_{2,0}$-norm penalized Graph Trend Filtering (GTF) model to estimate piecewise smooth graph signals that exhibit inhomogeneous levels of smoothness across the nodes. We prove that the proposed GTF model is simultaneously a k-means clustering on the signal over the nodes and a minimum graph cut on the edges of the graph, where the clustering and the cut share the same assignment matrix. We propose two methods to solve the proposed GTF model: a spectral decomposition method and a method based on simulated annealing. In the experiment on synthetic and real-world datasets, we show that the proposed GTF model has a better performances compared with existing approaches on the tasks of denoising, support recovery and semi-supervised classification. We also show that the proposed GTF model can be solved more efficiently than existing models for the dataset with a large edge set.
title Inhomogeneous graph trend filtering via a l2,0 cardinality penalty
topic Machine Learning
Social and Information Networks
65F50, 68U01, 68R01
G.1.6; G.1.10
url https://arxiv.org/abs/2304.05223