We describe a new approach for understanding the difficulty of designing efficient learning algorithms. We prove that the existence of an efficient learning algorithm for a circui...
A Bayesian framework for deformable pattern classification has been proposed in [1] with promising results for isolated handwritten character recognition. Its performance, however...
: This article describes a method for solving the geometric closure problem for simplified models of nucleic acid structures by using the constant bond lengths approximation. The r...
In this paper, we introduce localized homology, a theory for finding local geometric descriptions for topological attributes. Given a space and a cover of subspaces, we construct...
Abstract. Treewidth is a graph parameter with several interesting theoretical and practical applications. This survey reviews algorithmic results on determining the treewidth of a ...