Graphics based instructional software for decision tree analysis using Bayesian methodology

Abstract

We have designed and developed a graphics based software package “BAYES” to demonstrate the Bayesian methodology (before and after sampling) for solving decision problems under uncertainty involving two decision alternatives. The user is guided through a step by step procedure in developing the decision tree and evaluating the consequences at each node of the tree. Consequences are evaluated by computing the expected “pay offs” and the “opportunity losses”. Choosing the best decision after sampling involves computing various statistics such as EVSI (Expected Value of Sample Information), AMEOL (Average Minimum Expected Opportunity Loss), and ENGS (Expected Net Gain from Sampling). Detailed computations of these statistics are displayed only upon request. The users can thus focus more on the methodology and less on the computational aspects. This package has made a tremendous impact on classroom teaching for MBA and Executive Development Program participants in India and Singapore.