Safety in the Face of Unknown Unknowns: Algorithm Fusion in Data-driven Engineering Systems

Most current machine learning algorithms make highly confident yet incorrect classifications when faced with unexpected test samples from an unknown distribution different from training; such epistemic uncertainty (unknown unknowns) can have catastrophic safety implications. In this conceptual paper...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Proceedings of the ... IEEE International Conference on Acoustics, Speech and Signal Processing (1998) s. 8162 - 8166
Hlavní autoři: Kshetry, Nina, Varshney, Lav R.
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.05.2019
Témata:
AI safety
algorithm fusion
Data models
epistemic uncertainty
Knowledge engineering
metacognition
Safety
Sensors
Training
Uncertainty
Wastewater
wastewater treatment
ISSN:2379-190X
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Most current machine learning algorithms make highly confident yet incorrect classifications when faced with unexpected test samples from an unknown distribution different from training; such epistemic uncertainty (unknown unknowns) can have catastrophic safety implications. In this conceptual paper, we propose a method to leverage engineering science knowledge to control epistemic uncertainty and maintain decision safety. The basic idea is an algorithm fusion approach that combines data-driven learned models with physical system knowledge, to operate between the extremes of purely data-driven classifiers and purely engineering science rules. This facilitates the safe operation of data-driven engineering systems, such as wastewater treatment plants.
ISSN:2379-190X
DOI:10.1109/ICASSP.2019.8683392