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

Nina Kshetry, Lav R Varshney

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

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.

Original languageEnglish (US)
Title of host publication2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages8162-8166
Number of pages5
ISBN (Electronic)9781479981311
DOIs
StatePublished - May 2019
Event44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Brighton, United Kingdom
Duration: May 12 2019May 17 2019

Publication series

NameICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume2019-May
ISSN (Print)1520-6149

Conference

Conference44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019
CountryUnited Kingdom
CityBrighton
Period5/12/195/17/19

Fingerprint

Systems engineering
Fusion reactions
Wastewater treatment
Learning algorithms
Learning systems
Classifiers
Uncertainty

Keywords

  • AI safety
  • algorithm fusion
  • epistemic uncertainty
  • metacognition
  • wastewater treatment

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Kshetry, N., & Varshney, L. R. (2019). Safety in the Face of Unknown Unknowns: Algorithm Fusion in Data-driven Engineering Systems. In 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings (pp. 8162-8166). [8683392] (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 2019-May). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICASSP.2019.8683392

Safety in the Face of Unknown Unknowns : Algorithm Fusion in Data-driven Engineering Systems. / Kshetry, Nina; Varshney, Lav R.

2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. p. 8162-8166 8683392 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 2019-May).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kshetry, N & Varshney, LR 2019, Safety in the Face of Unknown Unknowns: Algorithm Fusion in Data-driven Engineering Systems. in 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings., 8683392, ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, vol. 2019-May, Institute of Electrical and Electronics Engineers Inc., pp. 8162-8166, 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019, Brighton, United Kingdom, 5/12/19. https://doi.org/10.1109/ICASSP.2019.8683392
Kshetry N, Varshney LR. Safety in the Face of Unknown Unknowns: Algorithm Fusion in Data-driven Engineering Systems. In 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2019. p. 8162-8166. 8683392. (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings). https://doi.org/10.1109/ICASSP.2019.8683392
Kshetry, Nina ; Varshney, Lav R. / Safety in the Face of Unknown Unknowns : Algorithm Fusion in Data-driven Engineering Systems. 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 8162-8166 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings).
@inproceedings{f1f5eb2ab5084bd29ff5130f45bad8e4,
title = "Safety in the Face of Unknown Unknowns: Algorithm Fusion in Data-driven Engineering Systems",
abstract = "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.",
keywords = "AI safety, algorithm fusion, epistemic uncertainty, metacognition, wastewater treatment",
author = "Nina Kshetry and Varshney, {Lav R}",
year = "2019",
month = "5",
doi = "10.1109/ICASSP.2019.8683392",
language = "English (US)",
series = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "8162--8166",
booktitle = "2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings",
address = "United States",

}

TY - GEN

T1 - Safety in the Face of Unknown Unknowns

T2 - Algorithm Fusion in Data-driven Engineering Systems

AU - Kshetry, Nina

AU - Varshney, Lav R

PY - 2019/5

Y1 - 2019/5

N2 - 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.

AB - 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.

KW - AI safety

KW - algorithm fusion

KW - epistemic uncertainty

KW - metacognition

KW - wastewater treatment

UR - http://www.scopus.com/inward/record.url?scp=85068958268&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068958268&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2019.8683392

DO - 10.1109/ICASSP.2019.8683392

M3 - Conference contribution

AN - SCOPUS:85068958268

T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

SP - 8162

EP - 8166

BT - 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -