Robustness of Deep ReLU Networks to Misclassification of High-Dimensional Data
Fuente:
arXiv
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908846250786816 |
|---|---|
| author | Kůrková, Věra |
| author_facet | Kůrková, Věra |
| contents | We present a theoretical study of the robustness of parameterized networks to random input perturbations. Specifically, we analyze local robustness at a given network input by quantifying the probability that a small additive random perturbation of the input leads to misclassification. For deep networks with rectified linear units, we derive lower bounds on local robustness in terms of the input dimensionality and the total number of network units. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_18674 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Robustness of Deep ReLU Networks to Misclassification of High-Dimensional Data Kůrková, Věra Machine Learning Neural and Evolutionary Computing 51, 60 G.0; G.3 We present a theoretical study of the robustness of parameterized networks to random input perturbations. Specifically, we analyze local robustness at a given network input by quantifying the probability that a small additive random perturbation of the input leads to misclassification. For deep networks with rectified linear units, we derive lower bounds on local robustness in terms of the input dimensionality and the total number of network units. |
| title | Robustness of Deep ReLU Networks to Misclassification of High-Dimensional Data |
| topic | Machine Learning Neural and Evolutionary Computing 51, 60 G.0; G.3 |
| url | https://arxiv.org/abs/2602.18674 |