Pattern variation is a major factor that affects the performance of recognition systems. In this paper, a novel manifold tangent modeling method called Discriminant Additive Tange...
: A new method for the characterization of molecules based on the model approach of molecular surfaces is presented. We use the topographical properties of the surface as well as t...
Abstract. In this paper we present a novel analysis of a random sampling approach for three clustering problems in metric spaces: k-median, min-sum kclustering, and balanced k-medi...
When correlating the samples with the corresponding class labels, canonical correlation analysis (CCA) can be used for supervised feature extraction and subsequent classification...
This work investigates how stochastic sampling jitter noise affects the result of system identification, and proposes a modification of known approaches to mitigate the effects of...