On the impact of downstream tasks on sampling and reconstructing noisy graph signals

Conference item

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.

Authors: Sripathmanathan, B; Dong, X; Bronstein, M

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: IEEE
• Host title: 2025 IEEE 10th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP)
• Pages: 111-115
• Publication date: 2026-03-12
• Event title: 10th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP 2025)
• Event location: Dominican Republic
• Event website: https://camsap25.ig.umons.ac.be/
• Event start date: 2025-12-14
• Event end date: 2025-12-17
• DOI: 10.1109/camsap66162.2025.11423962
• EISSN: 2994-8983
• ISSN: 2994-8991
• EISBN: 9798331526696
• ISBN: 9798331526702
• Language: English
• Keywords: adaptation models; graph convolutional networks; conferences; computational modeling; noise; reconstruction algorithms; sampling methods; signal reconstruction; noise measurement; noise level