LIFT: Learning Fault Trees from Observational Data

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Industries with safety-critical systems increasingly collect data on events occurring at the level of system components, thus capturing instances of system failure or malfunction. With data availability, it becomes possible to automatically learn a model describing the failure modes of the system, i.e., how the states of individual components combine to cause a system failure. We present LIFT, a machine learning method for static fault trees directly out of observational datasets. The fault trees model probabilistic causal chains of events ending in a global system failure. Our method makes use of the Mantel-Haenszel statistical test to narrow down possible causal relationships between events. We evaluate LIFT with synthetic case studies, show how its performance varies with the quality of the data, and discuss practical variants of LIFT.
Original languageEnglish
Title of host publicationQuantitative Evaluation of Systems
Subtitle of host publication15th International Conference, QEST 2018, Beijing, China, September 4-7, 2018, Proceedings
EditorsAnnabelle McIver, Andras Horvath
PublisherSpringer
ISBN (Electronic)978-3-319-99154-2
ISBN (Print)978-3-319-99153-5
Publication statusPublished - 2018
Event15th International Conference on Quantitative Evaluation of Systems, QEST 2018 - UCAS Campus, Beijing, China
Duration: 4 Sep 20187 Sep 2018
Conference number: 15
http://www.qest.org/qest2018/

Publication series

NameLecture Notes in Computer Science
Volume11024
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference15th International Conference on Quantitative Evaluation of Systems, QEST 2018
Abbreviated titleQEST 2018
CountryChina
CityBeijing
Period4/09/187/09/18
Internet address

Fingerprint

Statistical tests
Failure modes
Learning systems
Availability
Industry
Statistical Models

Cite this

Nauta, M., Bucur, D., & Stoelinga, M. (2018). LIFT: Learning Fault Trees from Observational Data. In A. McIver, & A. Horvath (Eds.), Quantitative Evaluation of Systems: 15th International Conference, QEST 2018, Beijing, China, September 4-7, 2018, Proceedings (Lecture Notes in Computer Science; Vol. 11024). Springer.
Nauta, Meike ; Bucur, Doina ; Stoelinga, Mariëlle. / LIFT : Learning Fault Trees from Observational Data. Quantitative Evaluation of Systems: 15th International Conference, QEST 2018, Beijing, China, September 4-7, 2018, Proceedings. editor / Annabelle McIver ; Andras Horvath. Springer, 2018. (Lecture Notes in Computer Science).
@inproceedings{8dc6970843db4e53a4c131b23b2fcbb4,
title = "LIFT: Learning Fault Trees from Observational Data",
abstract = "Industries with safety-critical systems increasingly collect data on events occurring at the level of system components, thus capturing instances of system failure or malfunction. With data availability, it becomes possible to automatically learn a model describing the failure modes of the system, i.e., how the states of individual components combine to cause a system failure. We present LIFT, a machine learning method for static fault trees directly out of observational datasets. The fault trees model probabilistic causal chains of events ending in a global system failure. Our method makes use of the Mantel-Haenszel statistical test to narrow down possible causal relationships between events. We evaluate LIFT with synthetic case studies, show how its performance varies with the quality of the data, and discuss practical variants of LIFT.",
author = "Meike Nauta and Doina Bucur and Mari{\"e}lle Stoelinga",
year = "2018",
language = "English",
isbn = "978-3-319-99153-5",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
editor = "Annabelle McIver and Andras Horvath",
booktitle = "Quantitative Evaluation of Systems",

}

Nauta, M, Bucur, D & Stoelinga, M 2018, LIFT: Learning Fault Trees from Observational Data. in A McIver & A Horvath (eds), Quantitative Evaluation of Systems: 15th International Conference, QEST 2018, Beijing, China, September 4-7, 2018, Proceedings. Lecture Notes in Computer Science, vol. 11024, Springer, 15th International Conference on Quantitative Evaluation of Systems, QEST 2018, Beijing, China, 4/09/18.

LIFT : Learning Fault Trees from Observational Data. / Nauta, Meike ; Bucur, Doina ; Stoelinga, Mariëlle.

Quantitative Evaluation of Systems: 15th International Conference, QEST 2018, Beijing, China, September 4-7, 2018, Proceedings. ed. / Annabelle McIver; Andras Horvath. Springer, 2018. (Lecture Notes in Computer Science; Vol. 11024).

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - LIFT

T2 - Learning Fault Trees from Observational Data

AU - Nauta, Meike

AU - Bucur, Doina

AU - Stoelinga, Mariëlle

PY - 2018

Y1 - 2018

N2 - Industries with safety-critical systems increasingly collect data on events occurring at the level of system components, thus capturing instances of system failure or malfunction. With data availability, it becomes possible to automatically learn a model describing the failure modes of the system, i.e., how the states of individual components combine to cause a system failure. We present LIFT, a machine learning method for static fault trees directly out of observational datasets. The fault trees model probabilistic causal chains of events ending in a global system failure. Our method makes use of the Mantel-Haenszel statistical test to narrow down possible causal relationships between events. We evaluate LIFT with synthetic case studies, show how its performance varies with the quality of the data, and discuss practical variants of LIFT.

AB - Industries with safety-critical systems increasingly collect data on events occurring at the level of system components, thus capturing instances of system failure or malfunction. With data availability, it becomes possible to automatically learn a model describing the failure modes of the system, i.e., how the states of individual components combine to cause a system failure. We present LIFT, a machine learning method for static fault trees directly out of observational datasets. The fault trees model probabilistic causal chains of events ending in a global system failure. Our method makes use of the Mantel-Haenszel statistical test to narrow down possible causal relationships between events. We evaluate LIFT with synthetic case studies, show how its performance varies with the quality of the data, and discuss practical variants of LIFT.

M3 - Conference contribution

SN - 978-3-319-99153-5

T3 - Lecture Notes in Computer Science

BT - Quantitative Evaluation of Systems

A2 - McIver, Annabelle

A2 - Horvath, Andras

PB - Springer

ER -

Nauta M, Bucur D, Stoelinga M. LIFT: Learning Fault Trees from Observational Data. In McIver A, Horvath A, editors, Quantitative Evaluation of Systems: 15th International Conference, QEST 2018, Beijing, China, September 4-7, 2018, Proceedings. Springer. 2018. (Lecture Notes in Computer Science).