-
Empowering the trustworthiness of ML-based critical systems through engineering activities
Abstract: This paper reviews the entire engineering process of trustworthy Machine Learning (ML) algorithms designed to equip critical systems with advanced analytics and decision functions. We start from the fundamental principles of ML and describe the core elements conditioning its trust, particularly through its design: namely domain specification, data engineering, design of the ML algorithms, their im… ▽ More
Submitted 30 September, 2022; originally announced September 2022.
Comments: This work has been supported by the French government under the "France 2030" program, as part of the SystemX Technological Research Institute Research Institute