Improved Bayesian Network-Based Risk Model and Its Application in Disaster Risk Assessment

Li, Ming; Mei, Hong; Zhang, Ren

doi:10.1007/s13753-018-0171-z

Cited by 26 publications

(18 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a tool for Environmental Risk Assessment, the Bayesian paradigm can synthesize different data types and account for the probabilities of different scenarios [33,68]. In addition, one of the significant benefits of the Bayesian approach is the ability to incorporate prior information [69,70] and extract qualitatively new knowledge from the available data with the help of machine learning techniques.…”

Section: Discussionmentioning

confidence: 99%

“…As the Bayesian belief networks allow, we used a combination of objective and subjective information. Where hard data were absent, we performed calculations based on specialists' estimations, as per the authors of [33]. A numerical value was attributed to all qualitative data in a format that is complementary to the quantitative data.…”

Section: Background Of the Methodology Used-bayesian Network For Influence Factorsmentioning

confidence: 99%

See 1 more Smart Citation

Measuring the Impact of Physical Geographical Factors on the Use of Coastal Zones Based on Bayesian Networks

et al. 2021

View full text Add to dashboard Cite

Coastal regions of the Baltic Sea are among the most intensively used worldwide, resulting in a need for a holistic management approach. Therefore, there is a need for strategies that even out the seasonality, which would ensure a better utilization of natural resources and infrastructure and improve the social and economic conditions. To assess the effectiveness of coastal zone planning processes concerning sustainable tourism and to identify and substantiate significant physical geographical factors impacting the sustainability of South Baltic seaside resorts, several data sets from previous studies were compiled. Seeking to improve the coastal zone’s ecological sustainability, economic efficiency, and social equality, a qualitative study (content analysis of planning documents) and a quantitative survey of tourists’ needs expressed on a social media platform and in the form of a survey, as well as long-term hydrometeorological data, were used. Furthermore, a Bayesian Network framework was used to combine knowledge from these different sources. We present an approach to identifying the social, economic, and environmental factors influencing the sustainability of coastal resorts. The results of this study may be used to advise local governments on a broad spectrum of Integrated Coastal Management matters: planning the development of the beaches and addressing the seasonality of use, directing investments to improve the quality of the beaches and protect them from storm erosion, and maintaining the sand quality and beach infrastructure. The lessons learned can be applied to further coastal zone management research by utilizing stakeholders and expert opinion in quantified current beliefs.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Background Of the Methodology Used-bayesian Network For Influence Factorsmentioning

confidence: 99%

Measuring the Impact of Physical Geographical Factors on the Use of Coastal Zones Based on Bayesian Networks

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The basic structure of a BN ( Fig 1A ) consists of nodes, each representing a variable, and arcs (arrows) which connect nodes in a causal relationship [ 22 ]. Underlying this intuitive and transparent structure are Bayes’ theorem and Bayesian statistics which allow for calculating conditional and joint probability distributions across many variables within the network [ 23 , 24 ]. The BN allows for both data and domain knowledge to be packaged into a single model that can yield insights about real-world probabilities, can be updated with new data or opinion over time and can tolerate missing information [ 13 , 21 , 22 ].…”

Section: Introductionmentioning

confidence: 99%

PI Prob: A risk prediction and clinical guidance system for evaluating patients with recurrent infections

et al. 2021

View full text Add to dashboard Cite

Background Primary immunodeficiency diseases represent an expanding set of heterogeneous conditions which are difficult to recognize clinically. Diagnostic rates outside of the newborn period have not changed appreciably. This concern underscores a need for novel methods of disease detection. Objective We built a Bayesian network to provide real-time risk assessment about primary immunodeficiency and to facilitate prescriptive analytics for initiating the most appropriate diagnostic work up. Our goal is to improve diagnostic rates for primary immunodeficiency and shorten time to diagnosis. We aimed to use readily available health record data and a small training dataset to prove utility in diagnosing patients with relatively rare features. Methods We extracted data from the Texas Children’s Hospital electronic health record on a large population of primary immunodeficiency patients (n = 1762) and appropriately-matched set of controls (n = 1698). From the cohorts, clinically relevant prior probabilities were calculated enabling construction of a Bayesian network probabilistic model(PI Prob). Our model was constructed with clinical-immunology domain expertise, trained on a balanced cohort of 100 cases-controls and validated on an unseen balanced cohort of 150 cases-controls. Performance was measured by area under the receiver operator characteristic curve (AUROC). We also compared our network performance to classic machine learning model performance on the same dataset. Results PI Prob was accurate in classifying immunodeficiency patients from controls (AUROC = 0.945; p<0.0001) at a risk threshold of ≥6%. Additionally, the model was 89% accurate for categorizing validation cohort members into appropriate International Union of Immunological Societies diagnostic categories. Our network outperformed 3 other machine learning models and provides superior transparency with a prescriptive output element. Conclusion Artificial intelligence methods can classify risk for primary immunodeficiency and guide management. PI Prob enables accurate, objective decision making about risk and guides the user towards the appropriate diagnostic evaluation for patients with recurrent infections. Probabilistic models can be trained with small datasets underscoring their utility for rare disease detection given appropriate domain expertise for feature selection and network construction.

show abstract

“…The Bayesian network (BN) is a graphical network of probabilistic reasoning, and it can analyze the uncertain relationship between variables in a complex network [ 17 ]. This method is widely used in risk analysis [ 18 , 19 , 20 , 21 ], risk assessment [ 22 , 23 , 24 , 25 ] and decision-making [ 26 , 27 ]. Zhu et al constructed a BN model of chemical terrorist attacks to conduct risk analysis, which provided theoretical support for the security prevention work of the risk management department [ 28 ].…”

Section: Introductionmentioning

confidence: 99%

Bayesian Network-Based Risk Analysis of Chemical Plant Explosion Accidents

Wang

Liu

2020

IJERPH

View full text Add to dashboard Cite

The chemical industry has made great contributions to the national economy, but frequent chemical plant explosion accidents (CPEAs) have also caused heavy property losses and casualties, as the CPEA is the result of interaction of many related risk factors, leading to uncertainty in the evolution of the accident. To systematically excavate and analyze the underlying causes of accidents, this paper first integrates emergency elements in the frame of orbit intersection theory and proposes 14 nodes to represent the evolution path of the accident. Then, combined with historical data and expert experience, a Bayesian network (BN) model of CPEAs was established. Through scenario analysis and sensitivity analysis, the interaction between factors and the impact of the factors on accident consequences was evaluated. It is found that the direct factors have the most obvious influence on the accident consequences, and the unsafe conditions contribute more than the unsafe behaviors. Furthermore, considering the factor chain, the management factors, especially safety education and training, are the key link of the accident that affects unsafe behaviors and unsafe conditions. Moreover, effective government emergency response has played a more prominent role in controlling environmental pollution. In addition, the complex network relationship between elements is presented in a sensitivity index matrix, and we extracted three important risk transmission paths from it. The research provides support for enterprises to formulate comprehensive safety production management strategies and control key factors in the risk transmission path to reduce CPEA risks.

show abstract

Improved Bayesian Network-Based Risk Model and Its Application in Disaster Risk Assessment

Cited by 26 publications

References 16 publications

Measuring the Impact of Physical Geographical Factors on the Use of Coastal Zones Based on Bayesian Networks

Measuring the Impact of Physical Geographical Factors on the Use of Coastal Zones Based on Bayesian Networks

PI Prob: A risk prediction and clinical guidance system for evaluating patients with recurrent infections

Bayesian Network-Based Risk Analysis of Chemical Plant Explosion Accidents

Contact Info

Product

Resources

About