Propositions as Subsets of the Data Space
Olaf Wolkenhauer · 2001
Concepts covered in this chapter include: A more general concept to represent data sampled from a system is that of a data space. System properties and behavior are reflected by clusters of data. Clusters may be interpreted as linear submodels of an overall nonlinear system. Clusters may also be interpreted as if-then rules relating properties of the variables that form the data space. Fuzzy clustering provides least-squares solutions to identify clusters, to partition the data space into clusters or classes. Fuzzy boundaries between clusters are differentiable functions and hence are computationally attractive. For many real-world problems a fuzzy partitioning of the underlying space is more realistic than ‘hard clustering’.