We o er a simple paradigm for tting models, parametric and non-parametric, to noisy data, which resolves some of the problems associated with classic MSE algorithms. This is done ...
In deep submicron technology, IR-drop and clock skew issues become more crucial to the functionality of chip. This paper presents a oorplan-based power and clock distribution meth...
er has outlined the potential of multiagent framework for decision support. From an abstract point of view, the concept of an agent has been used as modularization principle for th...
This paper reports a controlled study with statistical signi cance tests on ve text categorization methods: the Support Vector Machines (SVM), a k-Nearest Neighbor (kNN) classi er...
Software DSM systems su er from the high communication and coherence-induced overheads that limit performance. This paper introduces our e orts in reducing system overheads of a h...