China
Africa
Explore chapters and articles related to this topic
Decision Support System for Complex Systems Risk Assessment with Bayesian Networks
Published in Kim Phuc Tran, Machine Learning and Probabilistic Graphical Models for Decision Support Systems, 2023
This part is based on19 and its purpose is to establish a meta-model for risk management where risk is due to the occurence of an udesirable event. The Meta-Model for Risk management consists of an influence diagram (ID) which is an extension of Bayesian networks to allow evaluating alternative decisions and not only relationships as in BN. They are simple visual representations of a decision problem under uncertainty. Influence diagrams offer an intuitive way to identify and display the essential elements, including decisions, uncertainties, and preferences, and how they influence each other. It shows the dependencies among the variables more clearly than a decision tree. The subsequent paragraphs will present all the different variables that will be used by the ultimate influence diagram model in the established framework.
Dam portfolio risk management: What we learned from analyzing seven dams owned by the Regional Government of Extremadura (Spain)
Published in Jean-Pierre Tournier, Tony Bennett, Johanne Bibeau, Sustainable and Safe Dams Around the World, 2019
M. Setrakian-Melgonian, I. Escuder-Bueno, J.T. Castillo-Rodríguez, A. Morales-Torres, D. Simarro-Rey
Influence diagrams are compact conceptual representations of the logic of a system. An influence diagram is any representation including the relations between possible events, states of the environment, states of the system or subsystems, and consequences. Therefore, it allows integrating information on load events, dam performance and potential consequences downstream. When using an influence diagram for dam risk modelling, the diagram offers a visual representation of the risk model which represents the system, and each variable of the system is represented as a node and each relation as a connector. The nodes which integrate the model include the information that will be used to calculate probabilities of failure and potential consequences downstream.
Decision making under uncertainty
Published in Charles Yoe, Primer on Risk Analysis, 2019
Figure 7.12 begins to explore this problem with an influence diagram. An influence diagram is a graphical device showing the relationships among the decisions, the chance events, and the consequences. Influence diagrams use several conventions: Squares or rectangles depict decision nodesCircles or ovals depict chance nodesDiamonds depict consequence nodesLines or arcs connecting the nodes show the direction of influence.
A Creativity Support System Based on Causal Mapping
Published in Journal of Computer Information Systems, 2018
Cognitive mapping techniques [e.g., 25, 26, 49, 64, 72] that aim to extract subjective knowledge from individuals, and to represent them in a graphical way, have been used in developing creativity supporting systems. There are several different cognitive mapping techniques used in practice [e.g., 25, 26, 64]. Causal maps [e.g., 2, 1, 31, 33, 53] aim to represent the concepts and the causal relationships between them, allowing the modeling of complex chains of arguments [53]. Concept maps attempt to identify the relationships (which can be bi-directional) between the different concepts (which do not need to represent causality) and aim to generate ideas and to help in knowledge development by integrating old and new knowledge. Semantic maps are used to explore an idea by listing other ideas connected to it, and are helpful in obtaining a better understanding of an individual’s belief system; these are also known as mind maps [e.g., 11]. Influence diagrams are graphical models used to represent complex decision processes, based on uncertain information, allowing the development of probabilistic models (based on Bayesian networks) from expert knowledge [e.g., 5, 14, 23]. An in-depth discussion of cognitive mapping, and respective techniques, can be found in References [26] and [64].