On the Impact of Sample Size in Reconstructing Noisy Graph Signals: A Theoretical Characterisation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910032068608000 |
|---|---|
| author | Sripathmanathan, Baskaran Dong, Xiaowen Bronstein, Michael |
| author_facet | Sripathmanathan, Baskaran Dong, Xiaowen Bronstein, Michael |
| contents | Reconstructing a signal on a graph from noisy observations of a subset of the vertices is a fundamental problem in the field of graph signal processing. This paper investigates how sample size affects reconstruction error in the presence of noise via an in-depth theoretical analysis of the two most common reconstruction methods in the literature, least-squares reconstruction (LS) and graph-Laplacian regularised reconstruction (GLR). Our theorems show that at sufficiently low signal-to-noise ratios (SNRs), under these reconstruction methods we may simultaneously decrease sample size and decrease average reconstruction error. We further show that at sufficiently low SNRs, for LS reconstruction we have a $Λ$-shaped error curve and for GLR reconstruction, a sample size of $ O(\sqrt{N})$, where $N$ is the total number of vertices, results in lower reconstruction error than near full observation. We present thresholds on the SNRs, $τ$ and $τ_{GLR}$, below which the error is non-monotonic, and illustrate these theoretical results with experiments across multiple random graph models, sampling schemes and SNRs. These results demonstrate that any decision in sample-size choice has to be made in light of the noise levels in the data. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_16816 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | On the Impact of Sample Size in Reconstructing Noisy Graph Signals: A Theoretical Characterisation Sripathmanathan, Baskaran Dong, Xiaowen Bronstein, Michael Signal Processing Social and Information Networks Reconstructing a signal on a graph from noisy observations of a subset of the vertices is a fundamental problem in the field of graph signal processing. This paper investigates how sample size affects reconstruction error in the presence of noise via an in-depth theoretical analysis of the two most common reconstruction methods in the literature, least-squares reconstruction (LS) and graph-Laplacian regularised reconstruction (GLR). Our theorems show that at sufficiently low signal-to-noise ratios (SNRs), under these reconstruction methods we may simultaneously decrease sample size and decrease average reconstruction error. We further show that at sufficiently low SNRs, for LS reconstruction we have a $Λ$-shaped error curve and for GLR reconstruction, a sample size of $ O(\sqrt{N})$, where $N$ is the total number of vertices, results in lower reconstruction error than near full observation. We present thresholds on the SNRs, $τ$ and $τ_{GLR}$, below which the error is non-monotonic, and illustrate these theoretical results with experiments across multiple random graph models, sampling schemes and SNRs. These results demonstrate that any decision in sample-size choice has to be made in light of the noise levels in the data. |
| title | On the Impact of Sample Size in Reconstructing Noisy Graph Signals: A Theoretical Characterisation |
| topic | Signal Processing Social and Information Networks |
| url | https://arxiv.org/abs/2406.16816 |