On the Impact of Downstream Tasks on Sampling and Reconstructing Noisy Graph Signals

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Sripathmanathan, Baskaran, Dong, Xiaowen, Bronstein, Michael
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912585561931776
author Sripathmanathan, Baskaran
Dong, Xiaowen
Bronstein, Michael
author_facet Sripathmanathan, Baskaran
Dong, Xiaowen
Bronstein, Michael
contents We investigate graph signal reconstruction and sample selection for classification tasks. We present general theoretical characterisations of classification error applicable to multiple commonly used reconstruction methods, and compare that to the classical reconstruction error. We demonstrate the applicability of our results by using them to derive new optimal sampling methods for linearized graph convolutional networks, and show improvement over other graph signal processing based methods.
format Preprint
id arxiv_https___arxiv_org_abs_2509_10874
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On the Impact of Downstream Tasks on Sampling and Reconstructing Noisy Graph Signals
Sripathmanathan, Baskaran
Dong, Xiaowen
Bronstein, Michael
Signal Processing
Machine Learning
We investigate graph signal reconstruction and sample selection for classification tasks. We present general theoretical characterisations of classification error applicable to multiple commonly used reconstruction methods, and compare that to the classical reconstruction error. We demonstrate the applicability of our results by using them to derive new optimal sampling methods for linearized graph convolutional networks, and show improvement over other graph signal processing based methods.
title On the Impact of Downstream Tasks on Sampling and Reconstructing Noisy Graph Signals
topic Signal Processing
Machine Learning
url https://arxiv.org/abs/2509.10874