We introduce a third-party confidentiality-preserving machine learning scheme for soft-failure detection leveraging the robustness of the principal components algorithm to the changes in the rotation of the data axis. We demonstrate that random scrambling of the data is effective to hide sensitive telemetry information.
Confidentiality-preserving Machine Learning Scheme to Detect Soft-failures in Optical Communication Networks
Silva M. F.;Pacini A.;Sgambelluri A.;Paolucci F.;Valcarenghi L.
2022-01-01
Abstract
We introduce a third-party confidentiality-preserving machine learning scheme for soft-failure detection leveraging the robustness of the principal components algorithm to the changes in the rotation of the data axis. We demonstrate that random scrambling of the data is effective to hide sensitive telemetry information.File in questo prodotto:
| File | Dimensione | Formato | |
|---|---|---|---|
|
Confidentiality-preserving_Machine_Learning_Scheme_to_Detect_Soft-failures_in_Optical_Communication_Networks.pdf
non disponibili
Tipologia:
Documento in Pre-print/Submitted manuscript
Licenza:
Copyright dell'editore
Dimensione
771.49 kB
Formato
Adobe PDF
|
771.49 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
