Unsupervised partitioning of numerical attributes using fuzzy sets

The current paper presents an enhanced partitioning mechanism for numerical data. The efficiency of our method will be illustrated through a solid set of tests that have been performed. We have planned this partitioning phase as an initial step in a more complex algorithm to be further studied and implemented. The final goal is to use it for future decision making in automatic image annotation. Fuzzy Sets theory has been used as a base for our clustering algorithm and partitioning. We included this mechanism as a component of a framework we developed for image processing, more exactly for the image segmentation evaluation model we are building.