Automated problem detection for machine learning models

US11449798B2 · kind B2 · utility

Assignee

• AMAZON TECHNOLOGIES, INC. · US

Inventors

• Andrea Olgiati · Gilroy, US
• Maximiliano Maccanti · Redmond, US
• Arun Babu Nagarajan · Redmond, US
• Lakshmi Naarayanan Ramakrishnan · Redmond, US
• Urvashi Chowdhary · Seattle, US
• Gowda Dayananda Anjaneyapura Range · Redmond, US
• Zohar Karnin · Hoboken, US
• Laurence Louis Eric Rouesnel · New York, US
• Stefano Stefani · Issaquah, US
• Vladimir Zhukov · Moscow, RU

Key dates

• Filing date: Sep 30, 2019
• Grant date: Sep 20, 2022
• Expiry date: Jan 29, 2041

Classification

• Technology area (CPC G): Physics
• CPC primary: G06F11/0793
• WIPO field: Computer technology
• WIPO sector: Electrical engineering

Abstract

Methods, systems, and computer-readable media for automated problem detection for machine learning models are disclosed. A machine learning analysis system receives data associated with use of a machine learning model. The data was collected by a machine learning inference system and comprises input to the model or a plurality of inferences representing output of the machine learning model. The machine learning analysis system performs analysis of the data associated with the use of the machine learning model. The machine learning analysis system detects one or more problems associated with the use of the machine learning model based at least in part on the analysis. The machine learning analysis system initiates one or more remedial actions associated with the one or more problems associated with the use of the machine learning model.