Français
3. Measured characteristics of AI systems
3.1 Evaluation criteria
An evaluation criterion is a characteristic/property/attribute of a system that is considered to assess its value. The criteria most frequently used in AI are presented below.
-
Performance
The performance of an AI system corresponds to the accuracy of its outputs. As specified in the VIM , a system will be considered accurate if it is faithful (low variability of outputs in the presence of low variability of input data) and fair (for low...
The Ultimate Scientific and Technical Reference
This article is included in
Industry of the future
This offer includes:
Knowledge Base
Updated and enriched with articles validated by our scientific committees
Services
A set of exclusive tools to complement the resources
Practical Path
Operational and didactic, to guarantee the acquisition of transversal skills
Doc & Quiz
Interactive articles with quizzes, for constructive reading
Measured characteristics of AI systems
Evaluation campaigns: benchmarking, challenges and competitions
Bibliography
Standards and norms
- BIPM: International Vocabulary of Metrology – Fundamental and general concepts and associated terms (VIM) 3rd edition - JCGM 200 - 2012
https://www.bipm.org/utils/common/documents/jcgm/JCGM_200_2012.pdf
Events
METRICS project (2020-2023, funded by H2020) – Metrological evaluation and testing of robots in international competitions
The aim is to organize intelligent robot competitions in four fields: healthcare, agri-food, infrastructure inspection and maintenance, and agile production. In particular, the aim is to build a permanent structure bringing together all European skills to jointly provide a satisfactory...
The Ultimate Scientific and Technical Reference
