Skip to Main Content
This paper aims at classifying changed from unchanged pattern in multi-acquisition data using kernel based support vector data description (SVDD). Indeed, SVDD is a well known method allowing to map the data into a high dimensional features space where an hypersphere encloses most patterns belonging to the ”un-changed” class. In this work, we propose a new kernel function which combines the characteristics of basic kernel functions with new information about features distribution and then dependency between samples through copula theory that will be used for the first time to our knowledge in the SVDD framework. The effectiveness of the method is demonstrated on synthetic and real data sets.