scholarly journals Development of a hybrid Bayesian network model for predicting acute fish toxicity using multiple lines of evidence

2019 ◽  
Author(s):  
S. Jannicke Moe ◽  
Anders L. Madsen ◽  
Kristin A. Connors ◽  
Jane M. Rawlings ◽  
Scott E. Belanger ◽  
...  
Keyword(s):  
Bayesian Network ◽  
Expert Knowledge ◽  
Weight Of Evidence ◽  
List Type ◽  
Toxicity Data ◽  
Web Interface ◽  
Fish Embryo ◽  
Fish Toxicity ◽  
Embryo Toxicity ◽  
Lines Of Evidence

AbstractA Bayesian network was developed for predicting the acute toxicity intervals of chemical substances to fish, based on information on fish embryo toxicity (FET) in combination with other information. This model can support the use of FET data in a Weight-of-Evidence (WOE) approach for replacing the use of juvenile fish. The BN predicted correct toxicity intervals for 69%-80% of the tested substances. The model was most sensitive to components quantified by toxicity data, and least sensitive to components quantified by expert knowledge. The model is publicly available through a web interface. Further development of this model should include additional lines of evidence, refinement of the discretisation, and training with a larger dataset for weighting of the lines of evidence. A refined version of this model can be a useful tool for predicting acute fish toxicity, and a contribution to more quantitative WOE approaches for ecotoxicology and environmental assessment more generally.HighlightsA Bayesian network (BN) was developed to predict the toxicity of chemicals to fishThe BN uses fish embryo toxicity data in a quantitative weight-of-evidence approachThe BN integrates physical, chemical and toxicological properties of chemicalsCorrect toxicity intervals were predicted for 69-80% of test casesThe BN is publicly available for demonstration and testing through a web interface

scholarly journals Evaluation of a Bayesian Network for Strengthening the Weight of Evidence to Predict Acute Fish Toxicity from Fish Embryo Toxicity Data

2020 ◽  
Vol 16 (4) ◽  
pp. 452-460 ◽  
Author(s):  
Adam Lillicrap ◽  
S Jannicke Moe ◽  
Raoul Wolf ◽  
Kristin A Connors ◽  
Jane M Rawlings ◽  
...  
Keyword(s):  
Bayesian Network ◽  
Weight Of Evidence ◽  
Toxicity Data ◽  
Fish Embryo ◽  
Fish Toxicity ◽  
Embryo Toxicity

scholarly journals Development of a hybrid Bayesian network model for predicting acute fish toxicity using multiple lines of evidence

2020 ◽  
Vol 126 ◽  
pp. 104655 ◽  
Author(s):  
S. Jannicke Moe ◽  
Anders L. Madsen ◽  
Kristin A. Connors ◽  
Jane M. Rawlings ◽  
Scott E. Belanger ◽  
...  
Keyword(s):  
Bayesian Network ◽  
Network Model ◽  
Bayesian Network Model ◽  
Fish Toxicity ◽  
Lines Of Evidence

Use of fish embryo toxicity tests for the prediction of acute fish toxicity to chemicals

2013 ◽  
Vol 32 (8) ◽  
pp. 1768-1783 ◽  
Author(s):  
Scott E. Belanger ◽  
Jane M. Rawlings ◽  
Gregory J. Carr
Keyword(s):  
Toxicity Tests ◽  
Fish Embryo ◽  
Fish Toxicity ◽  
Embryo Toxicity

Traffic Incident Duration Estimation Based on a Dual-Learning Bayesian Network Model

2018 ◽  
Vol 2672 (45) ◽  
pp. 196-209 ◽  
Author(s):  
Haozhe Cong ◽  
Cong Chen ◽  
Pei-Sung Lin ◽  
Guohui Zhang ◽  
John Milton ◽  
...  
Keyword(s):  
Bayesian Network ◽  
Expert Knowledge ◽  
Transportation Systems ◽  
Practical Significance ◽  
Incident Management ◽  
Highway Traffic ◽  
Traffic Incidents ◽  
Traffic Incident ◽  
Incident Duration ◽  
Traffic Incident Management

Highway traffic incidents induce a significant loss of life, economy, and productivity through injuries and fatalities, extended travel time and delay, and excessive energy consumption and air pollution. Traffic emergency management during incident conditions is the core element of active traffic management, and it is of practical significance to accurately understand the duration time distribution for typical traffic incident types and the factors that influence incident duration. This study proposes a dual-learning Bayesian network (BN) model to estimate traffic incident duration and to examine the influence of heterogeneous factors on the length of duration based on expert knowledge of traffic incident management and highway incident data collected in Zhejiang Province, China. Fifteen variables related to three aspects of traffic incidents, including incident information, incident consequences, and rescue resources, were included in the analysis. The trained BN model achieves favorable performance in several areas, including classification accuracy, the receiver operating characteristic (ROC) curve, and the area under curve (AUC) value. A classification matrix, and significant variables and their heterogeneous influences are identified accordingly. The research findings from this study provide beneficial reference to the understanding of decision-making in traffic incident response and process, active traffic incident management, and intelligent transportation systems.

scholarly journals Bayesian Network Models with Decision Tree Analysis for Management of Childhood Malaria in Malawi

2020 ◽  
Author(s):  
Sanya B. Taneja ◽  
Gerald P. Douglas ◽  
Gregory F. Cooper ◽  
Marian G. Michaels ◽  
Marek J. Druzdzel ◽  
...  
Keyword(s):  
Decision Tree ◽  
Bayesian Network ◽  
Clinical Features ◽  
Predictive Models ◽  
Clinical Decision Making ◽  
Performance Metrics ◽  
Expert Knowledge ◽  
Clinical Decision ◽  
Automated Method ◽  
Childhood Malaria

Abstract Background: Malaria is a major cause of death in children under five years old in low- and middle-income countries such as Malawi. Accurate diagnosis and management of malaria can help reduce the global burden of childhood morbidity and mortality. Trained healthcare workers in rural health centers manage malaria with limited supplies of malarial diagnostic tests and drugs for treatment. A clinical decision support system that integrates predictive models to provide an accurate prediction of malaria based on clinical features could aid healthcare worker in judicious use of testing and treatment. We developed Bayesian network (BN) models to predict the probability of malaria from clinical features and an illustrative decision tree to model the decision to use or not use a malaria rapid diagnostic test (mRDT).Methods: We developed two BN models from data that were collected in a national survey of outpatient encounters of children in Malawi. The target diagnosis is taken as the result of mRDT. The first BN model was created manually with expert knowledge, and the second model was derived using an automated method followed by modifications guided by expert knowledge. The performance of the BN models was compared to other statistical models on a range of performance metrics. We developed a decision tree that integrates predictions from these predictive models with the costs of mRDT and a course of recommended treatment. Results: Compared to the logistic regression and random forest models, the BN models had similar accuracy of 64% but had higher sensitivity at the cost of lower specificity at the default threshold. Sensitivity analysis of the decision tree showed that at low (below 0.04) and high (above 0.4) probabilities of malaria in a child, the preferred decision that minimizes expected costs is not to perform mRDT.Conclusion: In resource-constrained settings, judicious use of mRDT is important. Predictive models in combination with decision analysis can provide personalized guidance on when to use mRDT in the management of childhood malaria. BN models can be efficiently derived from data to support such clinical decision making.

scholarly journals Probabilistic prediction of wildfire economic losses to housing in Cyprus using Bayesian network analysis

2017 ◽  
Vol 26 (1) ◽  
pp. 10 ◽  
Author(s):  
P. Papakosta ◽  
G. Xanthopoulos ◽  
D. Straub
Keyword(s):  
Bayesian Network ◽  
Prediction Models ◽  
Expert Knowledge ◽  
Fire Suppression ◽  
Fire Hazard ◽  
Economic Losses ◽  
Limited Attention ◽  
Probabilistic Prediction ◽  
Wildfire Risk ◽  
Wildfire Hazard

Loss prediction models are an important part of wildfire risk assessment, but have received only limited attention in the scientific literature. Such models can support decision-making on preventive measures targeting fuels or potential ignition sources, on fire suppression, on mitigation of consequences and on effective allocation of funds. This paper presents a probabilistic model for predicting wildfire housing loss at the mesoscale (1 km2) using Bayesian network (BN) analysis. The BN enables the construction of an integrated model based on causal relationships among the influencing parameters jointly with the associated uncertainties. Input data and models are gathered from literature and expert knowledge to overcome the lack of housing loss data in the study area. Numerical investigations are carried out with spatiotemporal datasets for the Mediterranean island of Cyprus. The BN is coupled with a geographic information system (GIS) and the resulting estimated house damages for a given fire hazard are shown in maps. The BN model can be attached to a wildfire hazard model to determine wildfire risk in a spatially explicit manner. The developed model is specific to areas with house characteristics similar to those found in Cyprus, but the general methodology is transferable to any other area, as well as other damages.

Fish embryo tests and acute fish toxicity tests are interchangeable in the application of the threshold approach

2019 ◽  
Vol 38 (3) ◽  
pp. 671-681 ◽  
Author(s):  
Jane M. Rawlings ◽  
Scott E. Belanger ◽  
Kristin A. Connors ◽  
Gregory J. Carr
Keyword(s):  
Toxicity Tests ◽  
Fish Embryo ◽  
Fish Toxicity ◽  
Threshold Approach
Sign in / Sign up

Export Citation Format

Share Document