ABOUT ONE METHOD FOR ASSESSING THE REPRESENTATIVENESS OF TESTING ARTIFICIAL INTELLIGENCE ALGORITHMS IN MEASUREMENT TOOLS

UDC 004.855
DOI: 10.36871/2618-9976.2023.10-2.001

Authors

Sergey V. Garbuk,
Ph.D., National Research University Higher School of Economics, Director of Scientific Projects, Technical Committee for Standardization TK164 "Artificial Intelligence", Chairman

Abstract

This paper proposes a machine learning approach to standardization in the field of intelligent change tools. Shows how machine learning methods can be used to compensate for random measurement errors. The essential operating factors and their variability are examined in detail, and it is also clearly shown how the creation of a regulatory and technical basis for testing intelligent measuring instruments and the corresponding conformity assessment system will ensure the development of AI algorithms that allow predicting measurement errors.

Keywords

Artificial intelligence, Machine learning, Significant operational factors, Training data set