Exam 1 Flashcards
What is a decision conference?
- A meeting where specialized decision support software is typically used
- A meeting where a model is built which includes factual data and value judgments elicited from the participants
What is even swaps?
- A simple method to help selecting the overall most attractive opinion in a multicriteria problem
- A method that uses dominance analysis between options
What is a ratio scale?
A ratios cale has an absolute zero and an arbitrary unit of measurement
What does the M-MACBETH decision support system?
Allows detecting dominates between options
A value scale generated by the decision support system M-MACBETH
- The value scale is obtained by solving a linear programming problem, which takes into account the judgments of the decision-maker
- Should be validated by the decision-maker as it is not unique
Which property should be observed by a single evaluation criterion in a multi criteria model?
- to be operational
- to e intelligible
- to be consensual
What descriptor of performance is best?
A direct descriptor of performance is preferable to a indirect descriptor and constructed descriptor of performance
Why should not ambiguous descriptors of performance be used?
They can confuse the decision maker
Name some information about PROBE ?
resource allocation to portfolios of projects
PROBE can deal with synergies involving more than two projects
common influence diagram?
starts with a decision node and then a chance node
what is the savage rule? (MiniMaxRegret)
The minimax regret approach is to minimize the worst-case regret.
what is conditional probability?
?
What is the expected value of information in a problem dominated by uncertainty?
- Expected value of information allows the decision-maker to know weather an expert is worth consulting for additional information
- Expected value of information always assumes a value between 0 and the expected value of perfect information
what is value of control?
?
Name some facts about the @RISKsoftware package!
- It allows modeling dependences among random variabels
- It produces cumulative risk profiles of the options in analysis, which allows comparing them
- @Risk requires a careful definition of the probability distributions of the input variables to produce adequate results
what is a key uncertainty?
- A key uncertainty is characterized by a strong potential impact on the focal issue of the scenarios exercise and a high degree of uncertainty.
- It would be relatively independent of other key uncertainties.
- Its parts and configurations should be reasonable, contrasting and challenging
what is a configuration of a key uncertainty?
?
what is a mega trend??
?
what is a wildcard?
Possible situations that happen sudden and with low probability, with high impact and surprise characteristics.
When a wildcard happen they deeply alter the structures being analysed
what is a weak signal?
?
What does the column 1:CT represent?
The sum of the configurations that the considered scenario has in common with other scenarious
What is the likelihood ratio of a test?
the ratio of the probability of finding it in patients with the disease with the probability of patients without the disease
where can the highest level of validity be found in the hierarchy of clinical evidence to support therapeutics?
Systematic review of clinical trials addressing the same question
which is the highest level concerning the categories of the sources for clinical information?
Systems
–> predictive systems based on databases of patients