Abstract--The particle swarm is an algorithm for finding optimal regions of complex search spaces through the interaction of individuals in a population of particles. Even though t...
In this paper, we propose a low-power approach to the design of embedded very long instruction word (VLIW) processor architectures based on the forwarding (or bypassing) hardware, ...
We analyze a separation procedure for Mixed-Integer Programs related to the work of Gomory and Johnson on interpolated subadditive functions. This approach has its roots in the Go...
In recent years, gradient-based LSTM recurrent neural networks (RNNs) solved many previously RNN-unlearnable tasks. Sometimes, however, gradient information is of little use for t...
—This paper presents a new framework for the completion of missing information based on local structures. It poses the task of completion as a global optimization problem with a ...