云计算环境下并行进化神经网络的设计研究

Research on Parallel Evolutionary Neural Network Design in Cloud Computing Environment

针对提高进化神经网络进化时效性,充分利用神经网络的训练数据,提出一种在云计算Hadoop平台环境下,使用进化算法对BP神经网络的权值和网络结构进行优化,通过分布并行计算,提高进化速度和效率。理论分析和实验结果表明,在数据量较大时,该方法能有效地提高神经网络计算精度。

In order to improve the evolutionary timeliness of evolutionary neural network and make full use of the training data of neural network,this paper proposes a method to optimize the weight and network structure of BP neural network by using evolutionary algorithm under the environment of Hadoop platform of cloud computing,meanwhile the evolutionary speed and efficiency is also improved through distributed parallel computing.The theoretical analysis and experimental results show that the method can effectively improve the accuracy of neural network when the amount of data is large.