We introduce perturbation kernels, a new class of similarity measure for information retrieval that casts word similarity in terms of multi-task learning. Perturbation kernels mode...
To align an IT system with an organization’s needs, it is necessary to understand the organization’s position within its environment as well as its internal configuration. In ...
Alain Wegmann, Gil Regev, Irina Rychkova, Lam-Son ...
— The increase in the demand for voice and data wireless services creates a need for a more efficient use of the available bandwidth. Current and future generations cellular sys...
Among the variety of alignment approaches (e.g., using machine learning, subsumption computation, formal concept analysis, etc.) similarity-based ones rely on a quantitative asses...
Abstract. We present a parameterized approximation scheme for distributed combinatorial optimization problems based on dynamic programming. The algorithm is a utility propagation m...