(An) interactive computer interview for probability encoding

Abstract

The uncertainty inherent in most decision problems usually requires the knowledge of one or more experts. Decision analysis uses a process of probability encoding to convert this specialized or general knowledge into probability distributions that represent judgment of these experts. This thesis represents the interview process that is considered to be the most effective method of encoding a probability distribution. This process is interactively programmed for personal and micro computer. We also construct the other programs for discrete approximations of a continuous probability distribution to facilitate computation. An illustration is given.