Method for processing of genetic algorithms equipped with individual neural networks using GPGPU units

摘要

method of processing genetic algorithms equipped with individual neural network using gpgpu units, is that in the process of processing is used wielordzeniowe measurement units gpgpu allowing calculation of high performance (hpc) and ubl systems iczeniowe equipped units gpgpu allocate each neural network reasonable cores.depending on the required degree of optimization process of their learning, and at the stage of testing the neural network computing systems equipped with units gpgpu apply operators macierzowe individual neural network and the functions wzbudzeu0144 to test input signals, in which one time processed many different neural network and multiple signal inputsthe crossing of parental chromosomes and introducing mutation and inversion use the functionality gpgpu processors, which allow for the parallel processing of multiple chromosomes at the same time, the decoding based on the unit wielordzeniowe gpgpu reason wish to generate many of parallel neural network at one time.