Method for decision and risk analysis in probabilistic and multiple criteria situations
OwnerMats Danielson, Love Ekenberginventors
Life of this patent
Estimating the term…
Abstract
A computer-implemented decision device and a family of methods and procedures are used for representing and analyzing imprecise, vague, and incomplete probabilistic decision and risk problems and/or when a criteria hierarchy is present. Decision makers can state the problems simply with the precision the decision-makers feel they have evidence for. Probabilities, values (utilities), criteria weights with belief distributions over them, and optimization procedures for fast evaluation of decision rules with respect to such statements are determined and analyzed.
Claims31
A computer system for decision making in a decision situation, the computer system comprising:
- an input unit for receiving data corresponding to weights and/or probabilities and values for a decision situation that can be represented in a generalized decision tree with criteria and probabilistic events in the same tree;
- a modeling unit for modeling the decision situation using at least one of a probabilistic decision model or a multi-criteria decision making model, where at least one of the probabilistic decision model or the multi-criteria decision making model is modeled in more than one level for generation of a decision frame, where the decision frame is a format storing information on interval event probabilities, interval criteria weights, interval utility values, and decision trees;
- a unit for evaluating the decision frame using input data fed through the input unit;
- and
- an output unit for outputting a decision prediction based on said decision frame evaluation.
An automated method of making decision predictions in a decision situation, the method comprising:
- receiving data corresponding to weights and/or probabilities and values for a decision situation that can be represented as a generalized decision tree with criteria and probabilistic events in the same tree;
- modeling the decision situation using at least one of a probabilistic decision model or a multi-criteria decision making model, where at least one of the probabilistic decision model or the multi-criteria decision making model is modeled in more than one level for generation of a decision frame, where a decision frame is a format storing information on interval event probabilities, interval criteria weights, interval utility values, and decision trees;
- evaluating the decision frame using input data fed through the input unit;
- generating a decision prediction based on said decision frame evluation;
- and
- outputting the decision prediction.
A computer software program product embodied in a computer readable medium including computer instructions for making decision predictions in a decision situation that when executed on a computer provides for the following steps to be executed:
- receiving data corresponding to weights and/or probabilities and values for a decision situation that can be represented as a generalized decision tree with criteria and probabilistic events in the same tree;
- modeling the decision situation using at least one of a probabilistic decision model or a multi-criteria decision making model, where at least one of the probabilistic decision model or the multi-criteria decision making model is modeled in more than one level for generation of a decision frame, wherein a decision frame is a format storing information on interval event probabilities, interval criteria weights, interval utility values, and decision trees;
- evaluating the decision frame using input data fed through the input unit;
- generating a decision prediction based on said decision frame evalution;
- and
- outputting the decision prediction.
Dependencies are inferred from the claim text.