论文信息 - Intrinsic and Parallel Performances of the OWE Neural Network Architecture

Intrinsic and Parallel Performances of the OWE Neural Network Architecture

The OWE (Orthogonal Weight Estimator) architecture is constituted of a main MLP in which the values of each weight is computed by another MLP (an OWE). The number of OWEs is equal to the number of weights of the main MLP. But the computation of each OWE is done independently. Therefore the training and relaxation phases can straightforward parallelized. We report the implementation of this architecture on an Intel Paragon parallel computer and the comparison with its implementation on a sequential computer.

Nicolas Pican

[1] Alexander Singer,et al. Implementations of artificial neural networks on the Connection Machine , 1990, Parallel Comput..

[2] Eugene J. H. Kerckhoffs,et al. Speeding up backpropagation training on a hypercube computer , 1992, Neurocomputing.

[3] Frédéric Alexandre,et al. A lateral contribution learning algorithm for multi MLP architecture , 1994, ESANN.

[4] Nicolas Pican. Synaptic efficiency modulations for context integration: The meta ODWE architecture , 1996, ESANN.

[5] Valmir Carneiro Barbosa. Massively parallel models of computation - distributed parallel processing in artificial intelligence and optimization , 1993, Ellis Horwood series in artificial intelligence.

[6] Davide Anguita,et al. An efficient implementation of BP on RISC-based workstations , 1994, Neurocomputing.

[7] Jean-François Mari,et al. HMMs and OWE neural network for continuous speech recognition , 1996, Proceeding of Fourth International Conference on Spoken Language Processing. ICSLP '96.