Abstract. We are interested in the relationship between learning efficiency and representation in the case of supervised neural networks for pattern classification trained by conti...
We present a performance analysis of an agent-based middleware system we have developed based on "reAgents," remotely executing agents that enhance the performance of cl...
We evaluate two broad classes of cognitive mechanisms that might support the learning of sequential patterns. According to the first, learning is based on the gradual accumulation...
A major difficulty in compiler development regards the proper modularization of concerns among the various compiler phases. The traditional object-oriented development paradigm ha...
The development of a highly configurable data management system is a challenging task, especially if it is to be implemented on an embedded system that provides limited resources...