Training set parallel implementations and analysis of backpropagation neural networks in a transputer array.
Foo, Shou King.
Date of Issue1994
School of Electrical and Electronic Engineering
This thesis presents a detailed study of the parallel implementations of backpropagation neural networks for reducing the training time. There are mainly two paradigms in the parallel implementations, namely, network based parallelism and training set parallelism. This study emphasises training set parallelism.
DRNTU::Engineering::Electrical and electronic engineering
NANYANG TECHNOLOGICAL UNIVERSITY