HOME > Research Results > Research Highlights, Machine Learning Quality Management Guideline

AIST:Research Highlights, Machine Learning Quality Management Guideline

Information Technology and Human Factors

Machine Learning Quality Management Guideline

OIWA Yutaka, Deputy Director, and Project Leader
Digital Architecture Research Center

Update: June 30, 2020

A guideline for quality management of machine learning systems

Machine Learning Quality Management Guideline” establishes a basis for quality goals for machine learning-based products/services and provides procedural guidance for realizing quality through development process management and system evaluations.

Machine Learning Quality Management Guideline

Quality management for machine learning systems is essential

Quality management is becoming indispensable for the deployment of AI-based systems to society. However, guaranteeing a certain level of quality for an AI system is challenging due to the nature of its internal structure. At present, there are no standards for AI system quality control and operation methods that cover the entire process of building AI systems. Companies developing and using AI systems are demanding guidelines as a basis for quality control methods.

2nd edition of the Guideline now available on the Internet

The second edition of the “Machine Learning Quality Management Guideline” is now available on the Digital Architecture Research Center and Cyber Physical Security Research Center websites and is now ready to use for quality evaluation and management of AI systems. The guideline's content is an outcome of joint efforts with experts from companies, universities, and other organizations involved with AI businesses.

Support for business activities and standardization of guidelines

Based on feedback from real business use, we plan to make regular revisions both in Japanese and English. We also plan to create reference guides for different industrial applications, provide a software development platform to assist evaluation work using this guideline, and develop various evaluation techniques. We also intend to contribute to international standardization based on the guideline contents.

Contact for inquiries related to this theme

Digital Architecture Research Center

OIWA Yutaka, Deputy Director

AIST Tokyo Waterfront, Annex (Bioscience and IT) 2-4-7 Aomi, Koto-ku, Tokyo 135-0064 Japan

E-mail: cpsec-inquiry-ml*aist.go.jp (Please convert "*" to "@".)

Web: https://www.cpsec.aist.go.jp/index_en.html

▲ ページトップへ