Anomaliekorrektur über Verschönerungsadverserale Angriffe

Details

Veröffentlichungs-Nr.

EP4769251

Anmeldetag

3. Dezember 2025

Veröffentlichung

1. Juli 2026

Rechtsraum

EP

IPC

G06N20/00

Abstract

A system and method of an anomaly detection and correction system for benevolent adversarial attacks. The method includes receiving a request to provide an input dataset to a machine learning (ML) model, trained using a training data distribution, to generate predictions based on input data. The method includes calculating, based on model traits of the model and the input dataset, an anomaly score indicating an accuracy of the model associated with generating a prediction output. The method includes adjusting, by a processing device based on the anomaly score, the input dataset to generate a modified input dataset. The method includes increasing a likelihood of an accurate prediction by the ML model when generating a prediction output by providing the modified input dataset to the model instead of the input dataset.

Anmelder

Firma

Red Hat, LLC

Land

🇺🇸 USA

Fachgebiete