Rule Classification Visualization of High-Dimensional Data
Frank Rehm, Frank Klawonn, Rudolf Kruse · elib (German Aerospace Center) · 2005
This paper presents an approach to visualize high-dimensional fuzzy classification rules and the corresponding classified data set in the plane. This enables the observer to check visually to which degree a feature vector is classified by a certain rule. Also misclassified feature vectors can be well spotted and conflicting or error-prone rules can be identified.