A Study on Decision Making Using Fuzzy Decision Trees

In operations research one often faces scenarios and phenomena exhibiting imprecision and uncertainty. Stochastic aspects of these scenarios are often accounted for by means of probabilistic models. In addition to these models, fuzzy augmentation to traditional decision analysis has been advocated as a possible method that takes into account imprecision in quantities available to the decision maker. Fuzzy methodology models these imprecise quantities as fuzzy numbers. Rules exist for computation of such quantities allowing for augmenting existing decision-making models. The objective of this thesis is to explore augmenting traditional decision trees with such fuzzy methods. Relevant definitions in the literature are given and an example decision tree is used to illustrate the augmentation process and further the results given by such fuzzy decision trees. Results given by the tree are then analyzed and compared to results given by traditional decision tree analysis. The contributions of the process are also discussed and further study is recommended.