Patent · US Active

Automated evaluation of machine learning models

US12242613B2 · kind B2 · utility

0Cited by

10References

25Claims

0Family size

Inventors

Mathieu Sinn · Dublin, IE
Beat Buesser · Dublin, IE
Ngoc Minh Tran · Dublin, IE
Killian Levacher · Dublin, IE
Hessel Tuinhof · Dublin, IE

Key dates

Filing date	Sep 30, 2020
Grant date	Mar 4, 2025
Priority date	—
Expiry date	Oct 15, 2041

Classification

Technology area (CPC G)Physics
CPC primaryG06F2221/2125
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

Various embodiments are provided for automated evaluation of machine learning models in a computing environment by one or more processors in a computing system. A level of robustness of a machine learning model against adversarial whitebox operations may be evaluated and determined by applying a data set used for testing the machine learning model, one or more adversarial operation objectives, an adversarial threat model, and a selected number of hyperparameters. Results from the adversarial operation may be analyzed and a modified machine learning model may be generated while performing the evaluating and determining.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.