This paper proposes a novel framework for automatic text categorization problem based on the kernel density classifier. The overall goal is to tackle two main issues in automatic ...
Dwi Sianto Mansjur, Ted S. Wada, Biing-Hwang Juang
Content fingerprinting provides a compact content-based representation of a multimedia document. An important application of fingerprinting is the identification of modified copie...
—We present a flexible interactive 3D morpho-kinematical modeling application for astrophysics. Compared to other systems, our application reduces the restrictions on the physic...
Wolfgang Steffen, Nicholas Koning, Stephan Wenger,...
Abstract We consider the problem of estimating the geometric deformation of an object, with respect to some reference observation on it. Existing solutions, set in the standard coo...
The problem of automatic feature selection/weighting in kernel methods is examined. We work on a formulation that optimizes both the weights of features and the parameters of the ...