Bayesian Solutions to a 2 x 2 Decision Matrix Using Interval Scaled Payoffs with an Application to Foreign Policy Decisions

摘要

Statistical Design Theory is applied to assessing objectively the uncertainty involved in foreign policy decisions. In an international conflict of interest situation, a protagonist is going to pursue. This problem is addressed by taking observations of the antagonist's behavior. These observations are interpreted as being associated with a specific course of action. Bayes formula is then used to update a conjectured a prior probability function to estimate the course of action being pursued by the antagonist. This updated (a posteriori) conditional probability function is then used to develop a decision rule to select an appropriate response. The decision rule is based on an ordering of possible outcomes, the values assigned to those outcomes, and greatest expected value of return.