Risks and their management are an integral part of decision making in the 21st century. A decision framework should allow groups of stakeholders to evaluate multiple-objective decisions and analyze trade-offs of various alternatives/scenarios/decision to be considered. However, formulating these alternatives, objectives, and trade-offs involves working with both facts (the uncertain state of the “world” one models) and values (what one deems important). Ideally, the facts and the values should be kept separate even when they seem completely entangled. This distinction is important both at a qualitative level, when building conceptual models, and at a quantitative level, when populating the conceptual models with numerical parameters. Moreover, at both levels, this distinction will directly help the search for, and use of appropriate resources.
In this Research Topic, we concentrate on modelling the uncertain state of the world and the way these models inform decision making and risk analysis. Nonetheless, perspectives on the larger decision framework scene are also welcome. Modelling uncertainty often requires the assessment of multiple, dependent uncertain quantities of interest. In addition to univariate distributions, interdependencies between these quantities or variables need to be modelled to properly understand the overall risk.
Various probabilistic dependence models could be used to represent multivariate distributions. The most advantageous models should be providing a transparent and efficient way to model complex relationships between observable and unobservable variables; they should be able to 1) integrate data from different sources with varying degrees of uncertainty, and 2) allow for the modelling of different dynamical processes under a single, statistically-robust framework.
This Research Topic intends to collect some of the most recent advances of multivariate probabilistic modelling, as embedded in specific risk analyses and decision problems. Contributions may include but not be restricted to research and review articles on:
• Foundational (theoretical) aspects
• Multivariate distributions
• Estimation and goodness-of-fit tests
• Risk measures
• Significant applications in science, engineering, ecology, etc.
• Probabilistic models’ quantification with, and/or in absence of data
• Computational methods and software
Risks and their management are an integral part of decision making in the 21st century. A decision framework should allow groups of stakeholders to evaluate multiple-objective decisions and analyze trade-offs of various alternatives/scenarios/decision to be considered. However, formulating these alternatives, objectives, and trade-offs involves working with both facts (the uncertain state of the “world” one models) and values (what one deems important). Ideally, the facts and the values should be kept separate even when they seem completely entangled. This distinction is important both at a qualitative level, when building conceptual models, and at a quantitative level, when populating the conceptual models with numerical parameters. Moreover, at both levels, this distinction will directly help the search for, and use of appropriate resources.
In this Research Topic, we concentrate on modelling the uncertain state of the world and the way these models inform decision making and risk analysis. Nonetheless, perspectives on the larger decision framework scene are also welcome. Modelling uncertainty often requires the assessment of multiple, dependent uncertain quantities of interest. In addition to univariate distributions, interdependencies between these quantities or variables need to be modelled to properly understand the overall risk.
Various probabilistic dependence models could be used to represent multivariate distributions. The most advantageous models should be providing a transparent and efficient way to model complex relationships between observable and unobservable variables; they should be able to 1) integrate data from different sources with varying degrees of uncertainty, and 2) allow for the modelling of different dynamical processes under a single, statistically-robust framework.
This Research Topic intends to collect some of the most recent advances of multivariate probabilistic modelling, as embedded in specific risk analyses and decision problems. Contributions may include but not be restricted to research and review articles on:
• Foundational (theoretical) aspects
• Multivariate distributions
• Estimation and goodness-of-fit tests
• Risk measures
• Significant applications in science, engineering, ecology, etc.
• Probabilistic models’ quantification with, and/or in absence of data
• Computational methods and software