Optimization of BP Neural Network Classifier Using Genetic Algorithm

Abstract

In this paper, a new BP neural network classifier was constructed and optimized by Genetic Algorithm, first, the BP neural network was improved by using genetic algorithm[2] to train the initial weights values of the BP neural network[3], second, a new classifier was constructed based on the new BP neural network optimized by Genetic Algorithm. Finally, data simulation experiment was taken and the result of data simulation with famous IRIS data shows that the new BP neural network classifier improved by the Genetic Algorithm has higher accuracy of classification and greater gradient of convergence than the BP Neural Network classifier which Proposed in literature [3].