1
|
Ramousse B, Mendoza-Lugo MA, Rongen G, Morales-Nápoles O. Elicitation of Rank Correlations with Probabilities of Concordance: Method and Application to Building Management. ENTROPY (BASEL, SWITZERLAND) 2024; 26:360. [PMID: 38785609 PMCID: PMC11120366 DOI: 10.3390/e26050360] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Received: 03/15/2024] [Revised: 04/19/2024] [Accepted: 04/23/2024] [Indexed: 05/25/2024]
Abstract
Constructing Bayesian networks (BN) for practical applications presents significant challenges, especially in domains with limited empirical data available. In such situations, field experts are often consulted to estimate the model's parameters, for instance, rank correlations in Gaussian copula-based Bayesian networks (GCBN). Because there is no consensus on a 'best' approach for eliciting these correlations, this paper proposes a framework that uses probabilities of concordance for assessing dependence, and the dependence calibration score to aggregate experts' judgments. To demonstrate the relevance of our approach, the latter is implemented to populate a GCBN intended to estimate the condition of air handling units' components-a key challenge in building asset management. While the elicitation of concordance probabilities was well received by the questionnaire respondents, the analysis of the results reveals notable disparities in the experts' ability to quantify uncertainty. Moreover, the application of the dependence calibration aggregation method was hindered by the absence of relevant seed variables, thus failing to evaluate the participants' field expertise. All in all, while the authors do not recommend to use the current model in practice, this study suggests that concordance probabilities should be further explored as an alternative approach for the elicitation of dependence.
Collapse
Affiliation(s)
- Benjamin Ramousse
- Department of Hydraulic Engineering, Delft University of Technology, 2628 CN Delft, The Netherlands
- Linesight, 75014 Paris, France
| | | | - Guus Rongen
- Department of Hydraulic Engineering, Delft University of Technology, 2628 CN Delft, The Netherlands
| | - Oswaldo Morales-Nápoles
- Department of Hydraulic Engineering, Delft University of Technology, 2628 CN Delft, The Netherlands
| |
Collapse
|
2
|
Hanea AM, Hilton Z, Knight B, P. Robinson A. Co-designing and building an expert-elicited non-parametric Bayesian network model: demonstrating a methodology using a Bonamia Ostreae spread risk case study. RISK ANALYSIS : AN OFFICIAL PUBLICATION OF THE SOCIETY FOR RISK ANALYSIS 2022; 42:1235-1254. [PMID: 35187670 PMCID: PMC9303608 DOI: 10.1111/risa.13904] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Indexed: 06/14/2023]
Abstract
The development and use of probabilistic models, particularly Bayesian networks (BN), to support risk-based decision making is well established. Striking an efficient balance between satisfying model complexity and ease of development requires continuous compromise. Codesign, wherein the structural content of the model is developed hand-in-hand with the experts who will be accountable for the parameter estimates, shows promise, as do so-called nonparametric Bayesian networks (NPBNs), which provide a light-touch approach to capturing complex relationships among nodes. We describe and demonstrate the process of codesigning, building, quantifying, and validating an NPBN model for emerging risks and the consequences of potential management decisions using structured expert judgment (SEJ). We develop a case study of the local spread of a marine pathogen, namely, Bonamia ostreae. The BN was developed through a series of semistructured workshops that incorporated extensive feedback from many experts. The model was then quantified with a combination of field and expert-elicited data. The IDEA protocol for SEJ was used in its hybrid (remote and face-to-face) form to elicit information about more than 100 parameters. This article focuses on the modeling and quantification process, the methodological challenges, and the way these were addressed.
Collapse
Affiliation(s)
- Anca M. Hanea
- Centre of Excellence for Biosecurity Risk AnalysisUniversity of MelbourneParkvilleVictoriaAustralia
| | | | | | - Andrew P. Robinson
- Centre of Excellence for Biosecurity Risk AnalysisUniversity of MelbourneParkvilleVictoriaAustralia
| |
Collapse
|
3
|
Falconer JR, Frank E, Polaschek DLL, Joshi C. Methods for Eliciting Informative Prior Distributions: A Critical Review. DECISION ANALYSIS 2022. [DOI: 10.1287/deca.2022.0451] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022]
Abstract
Eliciting informative prior distributions for Bayesian inference can often be complex and challenging. Although popular methods rely on asking experts probability-based questions to quantify uncertainty, these methods are not without their drawbacks, and many alternative elicitation methods exist. This paper explores methods for eliciting informative priors categorized by type and briefly discusses their strengths and limitations. Most of the review literature in this field focuses on a particular type of elicitation approach. The primary aim of this work, however, is to provide a more complete yet macro view of the state of the art by highlighting new (and old) approaches in one clear easy-to-read article. Two representative applications are used throughout to explore the suitability, or lack thereof, of the existing methods, one of which highlights a challenge that has not been addressed in the literature yet. We identify some of the gaps in the present work and discuss directions for future research.
Collapse
Affiliation(s)
- Julia R. Falconer
- Department of Mathematics, University of Waikato, Hamilton 3216, New Zealand
- New Zealand Institute of Security and Crime Science, University of Waikato, Hamilton 3216, New Zealand
| | - Eibe Frank
- Department of Computer Science, University of Waikato, Hamilton 3216, New Zealand
| | - Devon L. L. Polaschek
- New Zealand Institute of Security and Crime Science, University of Waikato, Hamilton 3216, New Zealand
- School of Psychology, University of Waikato, Hamilton 3216, New Zealand
| | - Chaitanya Joshi
- Department of Mathematics, University of Waikato, Hamilton 3216, New Zealand
- New Zealand Institute of Security and Crime Science, University of Waikato, Hamilton 3216, New Zealand
| |
Collapse
|
4
|
Swallow B, Birrell P, Blake J, Burgman M, Challenor P, Coffeng LE, Dawid P, De Angelis D, Goldstein M, Hemming V, Marion G, McKinley TJ, Overton CE, Panovska-Griffiths J, Pellis L, Probert W, Shea K, Villela D, Vernon I. Challenges in estimation, uncertainty quantification and elicitation for pandemic modelling. Epidemics 2022; 38:100547. [PMID: 35180542 PMCID: PMC7612598 DOI: 10.1016/j.epidem.2022.100547] [Citation(s) in RCA: 9] [Impact Index Per Article: 4.5] [Reference Citation Analysis] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/07/2021] [Revised: 12/22/2021] [Accepted: 02/09/2022] [Indexed: 12/15/2022] Open
Affiliation(s)
- Ben Swallow
- School of Mathematics and Statistics, University of Glasgow, Glasgow, UK; Scottish COVID-19 Response Consortium, UK.
| | - Paul Birrell
- Analytics & Data Science, UKHSA, UK; MRC Biostatistics Unit, University of Cambridge, Cambridge, UK
| | - Joshua Blake
- MRC Biostatistics Unit, University of Cambridge, Cambridge, UK
| | - Mark Burgman
- Centre for Environmental Policy, Imperial College London, London, UK
| | - Peter Challenor
- The Alan Turing Institute, London, UK; College of Engineering, Mathematics and Physical Sciences, University of Exeter, Exeter, UK
| | - Luc E Coffeng
- Department of Public Health, Erasmus MC, University Medical Center Rotterdam, Rotterdam, The Netherlands
| | - Philip Dawid
- Statistical Laboratory, University of Cambridge, Cambridge, UK
| | - Daniela De Angelis
- MRC Biostatistics Unit, University of Cambridge, Cambridge, UK; Joint UNIversities Pandemic and Epidemiological Research, UK
| | - Michael Goldstein
- Department of Mathematical Sciences, Durham University, Stockton Road, Durham, UK
| | - Victoria Hemming
- Department of Forest and Conservation Sciences, University of British Columbia, Vancouver, Canada
| | - Glenn Marion
- Scottish COVID-19 Response Consortium, UK; Biomathematics and Statistics Scotland, Edinburgh, UK
| | - Trevelyan J McKinley
- College of Medicine and Health, University of Exeter, Exeter, UK; Joint UNIversities Pandemic and Epidemiological Research, UK
| | - Christopher E Overton
- Department of Mathematics, University of Manchester, Manchester, UK; Clinical Data Science Unit, Manchester University NHS Foundation Trust, Manchester, UK; Joint UNIversities Pandemic and Epidemiological Research, UK
| | - Jasmina Panovska-Griffiths
- The Big Data Institute, University of Oxford, Oxford, UK; The Queen's College, University of Oxford, Oxford, UK
| | - Lorenzo Pellis
- Department of Mathematics, University of Manchester, Manchester, UK; Joint UNIversities Pandemic and Epidemiological Research, UK; The Alan Turing Institute, London, UK
| | - Will Probert
- The Big Data Institute, University of Oxford, Oxford, UK
| | - Katriona Shea
- Department of Biology and Centre for Infectious Disease Dynamics, The Pennsylvania State University, PA 16802, USA
| | - Daniel Villela
- Program of Scientific Computing, Fundação Oswaldo Cruz, Rio de Janeiro, Brazil
| | - Ian Vernon
- Department of Mathematical Sciences, Durham University, Stockton Road, Durham, UK
| |
Collapse
|
5
|
Adamchick J, Pérez Aguirreburualde MS, Perez AM, O'Brien MK. One Coin, Two Sides: Eliciting Expert Knowledge From Training Participants in a Capacity-Building Program for Veterinary Professionals. Front Vet Sci 2021; 8:729159. [PMID: 34760954 PMCID: PMC8573137 DOI: 10.3389/fvets.2021.729159] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/22/2021] [Accepted: 09/27/2021] [Indexed: 11/28/2022] Open
Abstract
Scientific research may include the elicitation of judgment from non-academic subject-matter experts in order to improve the quality and/or impact of research studies. Elicitation of expert knowledge or judgment is used when data are missing, incomplete, or not representative for the specific setting and processes being studied. Rigorous methods are crucial to ensure robust study results, and yet the quality of the elicitation can be affected by a number of practical constraints, including the understanding that subject-matter experts have of the elicitation process itself. In this paper, we present a case of expert elicitation embedded within an extended training course for veterinary professionals as an example of overcoming these constraints. The coupling of the two activities enabled extended opportunities for training and a relationship of mutual respect to be the foundation for the elicitation process. In addition, the participatory research activities reinforced knowledge synthesis objectives of the educational program. Finally, the synergy between the two concurrent objectives may produce benefits which transcend either independent activity: solutions and ideas built by local professionals, evolving collaborative research and training approaches, and a network of diverse academic and practicing professionals. This approach has the versatility to be adapted to many training and research opportunities.
Collapse
Affiliation(s)
- Julie Adamchick
- Department of Veterinary Population Medicine, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN, United States
| | - María Sol Pérez Aguirreburualde
- Center for Animal Health and Food Safety, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN, United States
| | - Andres M. Perez
- Department of Veterinary Population Medicine, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN, United States
| | - Mary Katherine O'Brien
- Center for Animal Health and Food Safety, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN, United States
| |
Collapse
|
6
|
Adamchick J, Rich KM, Perez AM. Self-Reporting of Risk Pathways and Parameter Values for Foot-and-Mouth Disease in Slaughter Cattle from Alternative Production Systems by Kenyan and Ugandan Veterinarians. Viruses 2021; 13:v13112112. [PMID: 34834919 PMCID: PMC8621966 DOI: 10.3390/v13112112] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/21/2021] [Revised: 10/09/2021] [Accepted: 10/13/2021] [Indexed: 01/07/2023] Open
Abstract
Countries in which foot-and-mouth disease (FMD) is endemic may face bans on the export of FMD-susceptible livestock and products because of the associated risk for transmission of FMD virus. Risk assessment is an essential tool for demonstrating the fitness of one’s goods for the international marketplace and for improving animal health. However, it is difficult to obtain the necessary data for such risk assessments in many countries where FMD is present. This study bridged the gaps of traditional participatory and expert elicitation approaches by partnering with veterinarians from the National Veterinary Services of Kenya (n = 13) and Uganda (n = 10) enrolled in an extended capacity-building program to systematically collect rich, local knowledge in a format appropriate for formal quantitative analysis. Participants mapped risk pathways and quantified variables that determine the risk of infection among cattle at slaughter originating from each of four beef production systems in each country. Findings highlighted that risk processes differ between management systems, that disease and sale are not always independent events, and that events on the risk pathway are influenced by the actions and motivations of value chain actors. The results provide necessary information for evaluating the risk of FMD among cattle pre-harvest in Kenya and Uganda and provide a framework for similar evaluation in other endemic settings.
Collapse
Affiliation(s)
- Julie Adamchick
- Department of Veterinary Population Medicine, College of Veterinary Medicine, University of Minnesota, Minneapolis, MN 55108, USA;
- Correspondence:
| | - Karl M. Rich
- Department of Agricultural Economics, Ferguson College of Agriculture, Oklahoma State University, Stillwater, OK 74078, USA;
| | - Andres M. Perez
- Department of Veterinary Population Medicine, College of Veterinary Medicine, University of Minnesota, Minneapolis, MN 55108, USA;
| |
Collapse
|