For character recognition in document analysis, some classes are closely overlapped but are not necessarily to be separated before contextual information is exploited. For classifi...
A wide variety of texture feature extraction methods have been proposed for texture based image classification and segmentation. These methods are typically evaluated over windows...
In this paper we present an analysis of the minimal hardware precision required to implement Support Vector Machine (SVM) classification within a Logarithmic Number System archite...
Faisal M. Khan, Mark G. Arnold, William M. Potteng...
A training data selection method for multi-class data is proposed. This method can be used for multilayer neural networks (MLNN). The MLNN can be applied to pattern classification...
We present a Generalized Lotka-Volterra (GLV) based approach for modeling and simulation of supervised inductive learning, and construction of an efficient classification algorith...
Karen Hovsepian, Peter Anselmo, Subhasish Mazumdar