Influences of variables on ship collision probability in a Bayesian belief network model

Maria Hänninen; Pentti Kujala

doi:10.1016/j.ress.2012.02.008

Influences of variables on ship collision probability in a Bayesian belief network model

Maria Hänninen^*
, Pentti Kujala

^*Corresponding author for this work

Not published at VTT

Aalto University

Research output: Contribution to journal › Article › Scientific › peer-review

201 Citations (Scopus)

Abstract

The influences of the variables in a Bayesian belief network model for estimating the role of human factors on ship collision probability in the Gulf of Finland are studied for discovering the variables with the largest influences and for examining the validity of the network. The change in the so-called causation probability is examined while observing each state of the network variables and by utilizing sensitivity and mutual information analyses. Changing course in an encounter situation is the most influential variable in the model, followed by variables such as the Officer of the Watchs action, situation assessment, danger detection, personal condition and incapacitation. The least influential variables are the other distractions on bridge, the bridge view, maintenance routines and the officers fatigue. In general, the methods are found to agree on the order of the model variables although some disagreements arise due to slightly dissimilar approaches to the concept of variable influence. The relative values and the ranking of variables based on the values are discovered to be more valuable than the actual numerical values themselves. Although the most influential variables seem to be plausible, there are some discrepancies between the indicated influences in the model and literature. Thus, improvements are suggested to the network.

Original language	English
Pages (from-to)	27-40
Number of pages	14
Journal	Reliability Engineering and System Safety
Volume	102
DOIs	https://doi.org/10.1016/j.ress.2012.02.008
Publication status	Published - Jun 2012
MoE publication type	A1 Journal article-refereed

Funding

The study was conducted as a part of SAFGOF and CAFE projects, financed by the European Union – European Regional Development Fund – Regional Councils of Kymenlaakso and Päijät-Häme, the City of Kotka, Kotka-Hamina regional development company Cursor Ltd., Kotka Maritime Research Association Merikotka and the following members of the Kotka Maritime Research Centre Corporate Group: Port of HaminaKotka, Aker Arctic Technology Inc. and Arctia Shipping Ltd. The authors wish to express their gratitude to the funders. D.Sc., Chief Mate Jakub Montewka and M.Soc.Sc. Jenni Storgård are thanked for providing their expertise on determining some of the dependencies and probability values of the analyzed model. Other colleagues at the Aalto University Department of Applied Mechanics Marine Technology research group and at Kotka Maritime Research Centre are thanked for given valuable comments on the manuscript.

Keywords

Bayesian networks
Causation probability
Maritime accidents
Mutual information
Sensitivity analysis

Access to Document

10.1016/j.ress.2012.02.008

Cite this

@article{920ed05b98ce406abb9182fe0e574af5,

title = "Influences of variables on ship collision probability in a Bayesian belief network model",

abstract = "The influences of the variables in a Bayesian belief network model for estimating the role of human factors on ship collision probability in the Gulf of Finland are studied for discovering the variables with the largest influences and for examining the validity of the network. The change in the so-called causation probability is examined while observing each state of the network variables and by utilizing sensitivity and mutual information analyses. Changing course in an encounter situation is the most influential variable in the model, followed by variables such as the Officer of the Watchs action, situation assessment, danger detection, personal condition and incapacitation. The least influential variables are the other distractions on bridge, the bridge view, maintenance routines and the officers fatigue. In general, the methods are found to agree on the order of the model variables although some disagreements arise due to slightly dissimilar approaches to the concept of variable influence. The relative values and the ranking of variables based on the values are discovered to be more valuable than the actual numerical values themselves. Although the most influential variables seem to be plausible, there are some discrepancies between the indicated influences in the model and literature. Thus, improvements are suggested to the network.",

keywords = "Bayesian networks, Causation probability, Maritime accidents, Mutual information, Sensitivity analysis",

author = "Maria H{\"a}nninen and Pentti Kujala",

year = "2012",

month = jun,

doi = "10.1016/j.ress.2012.02.008",

language = "English",

volume = "102",

pages = "27--40",

journal = "Reliability Engineering and System Safety",

issn = "0951-8320",

publisher = "Elsevier",

}

TY - JOUR

T1 - Influences of variables on ship collision probability in a Bayesian belief network model

AU - Hänninen, Maria

AU - Kujala, Pentti

PY - 2012/6

Y1 - 2012/6

N2 - The influences of the variables in a Bayesian belief network model for estimating the role of human factors on ship collision probability in the Gulf of Finland are studied for discovering the variables with the largest influences and for examining the validity of the network. The change in the so-called causation probability is examined while observing each state of the network variables and by utilizing sensitivity and mutual information analyses. Changing course in an encounter situation is the most influential variable in the model, followed by variables such as the Officer of the Watchs action, situation assessment, danger detection, personal condition and incapacitation. The least influential variables are the other distractions on bridge, the bridge view, maintenance routines and the officers fatigue. In general, the methods are found to agree on the order of the model variables although some disagreements arise due to slightly dissimilar approaches to the concept of variable influence. The relative values and the ranking of variables based on the values are discovered to be more valuable than the actual numerical values themselves. Although the most influential variables seem to be plausible, there are some discrepancies between the indicated influences in the model and literature. Thus, improvements are suggested to the network.

AB - The influences of the variables in a Bayesian belief network model for estimating the role of human factors on ship collision probability in the Gulf of Finland are studied for discovering the variables with the largest influences and for examining the validity of the network. The change in the so-called causation probability is examined while observing each state of the network variables and by utilizing sensitivity and mutual information analyses. Changing course in an encounter situation is the most influential variable in the model, followed by variables such as the Officer of the Watchs action, situation assessment, danger detection, personal condition and incapacitation. The least influential variables are the other distractions on bridge, the bridge view, maintenance routines and the officers fatigue. In general, the methods are found to agree on the order of the model variables although some disagreements arise due to slightly dissimilar approaches to the concept of variable influence. The relative values and the ranking of variables based on the values are discovered to be more valuable than the actual numerical values themselves. Although the most influential variables seem to be plausible, there are some discrepancies between the indicated influences in the model and literature. Thus, improvements are suggested to the network.

KW - Bayesian networks

KW - Causation probability

KW - Maritime accidents

KW - Mutual information

KW - Sensitivity analysis

UR - https://www.scopus.com/pages/publications/84858134422

U2 - 10.1016/j.ress.2012.02.008

DO - 10.1016/j.ress.2012.02.008

M3 - Article

AN - SCOPUS:84858134422

SN - 0951-8320

VL - 102

SP - 27

EP - 40

JO - Reliability Engineering and System Safety

JF - Reliability Engineering and System Safety

ER -

Influences of variables on ship collision probability in a Bayesian belief network model

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this