Mid Sweden University

This paper generalizes a method for handling decisions, when vague and numerically imprecise information prevails, thus extending the Delta approach into a method for handling multi-level decision trees. The straight-forward way of doing this is by using a multi-linear solver; however, this is very demanding from a computational point of view. The proposed solution is to use either backward induction or recursively collapse the multi-level decision tree into a single-level tree and thus mapping it to a bilinear problem. Restricting the format of these types of decision problems, they can be solved with reasonable computational efforts.

The nature of much information available to decision makers is vague and imprecise, be it information for human managers in organisations or for process agents in a distributed computer environment. Several models for handling vague and imprecise information in decision situations have been suggested. In particular, various interval methods have prevailed, i.e. methods based on interval estimates of probabilities and, in some cases, interval utility estimates. Even if these approaches in general are well founded, little has been done to take into consideration the evaluation perspective and, in particular, computational aspects and implementation issues. The purpose of this paper is to demonstrate a tool for handling imprecise information in decision situations. The tool is an implementation of our earlier research focussing on finding fast algorithms for solving bilinear systems of equations together with a graphical user interface supporting the interpretation of evaluations of imprecise data.

In investment decision-making in organizations, large values can be at stake. The need for a structured process is evident, not least when several people or interests are involved in the process. We advocate that the use of modern computational decision analysis can improve such investment processes by improving visibility and only requiring reasonably precise input data. The applicability of a structured decision analysis to corporate decision-making is demonstrated in a case study at SCA Transforest, a subdivision to SCA. The decision problem consists of whether a new system for logistic control should be implemented or not. The background information was collected through interviews and the structuring and analysis of the problem was performed using the tool DecideIT, designed for handling situations where uncertainties in input data prevail. The result of the analysis points out a reasonable action, but also shows which aspects are crucial to consider for a reliable result.

This article describes the application of computational decision analytic techniques for a national policy decision. It constitutes an example of the increasing use of modern computational decision methods to assist in decision-making in society. An integrated flood catastrophe model is presented as well as some results of a case study made in the Upper Tisza region in north-eastern Hungary, viz. the Palad-Csecsei basin. Background data was provided through the Hungarian Academy of Sciences and complemented by interviews with different stakeholders in the region. Based upon these data, where a large degree of uncertainty is prevailing, we demonstrate how an implementation of a simulation and decision analytical model can provide insights into the effects of imposing different policy options for a flood risk management program in the region. We focus herein primarily on general options for designing a public/private insurance and reinsurance system for Hungary. It should, however, be emphasized that the main purpose of this article is not to provide any definite recommendations, but rather to present a methodology for handling a set of policy packages with the aim of gaining a consensus among stakeholders.

We present a decision tree evaluation method integrated with a common framework for analyzing multi-attribute decisions under risk, where information is numerically imprecise. The approach extends the use of additive and multiplicative utility functions for supporting evaluation of imprecise statements, relaxing requirements for precise estimates of decision parameters. Information is modeled in convex sets of utility and probability measures restricted by closed intervals. Evaluation is done relative to a set of rules, generalizing the concept of admissibility, computationally handled through optimization of aggregated utility functions. Pros and cons of two approaches, and tradeoffs in selecting a utility function, are discussed.