Applying Fuzzy Clustering for a Better Representation of the Total Scenario Space

In scenario analysis we are looking for some consistent, but quite different pictures of the future concerning a given problem. Not just the pictures themselves but also the way from today to a forward-looking situation shall be explained. Agglomerative methods in classification are already used to find the right number of final scenarios and to determine the final scenarios themselves. Here we examine, whether these questions can better be answered with the help of fuzzy-c-means cluster algorithm or whether fuzzy classification results can support interpretation and decision making in that context. Two different approaches will be presented.