Parallel Implementation of a Neural Network Learning Algorithm
S. Volokitin · International Journal of Computer Applications · 2014
This paper describes parallel implementation of an artificial neural network training algorithm and its effectiveness when applied to performing cryptographic functions.As a cryptographic function a permutations have been used because of its prevalence in complex cryptographic functions such as block ciphers.In order to enhance performance of artificial neural network training algorithm a method of backward propagation of errors has been parallelized.