CAJ | 학술논문

BP 神经网络已经成为应用最为广泛的神经网络模型之一.而人工神经网络是对人脑真正神经工作的简化生物模型.为了加深对神经网络的理解,利用推导公式来详细分析其最后的输出值和误差.这里旨在阐述用C语言实现BP神经网络基本模型,在BP神经网络的初始化函数中采用了归一化处理的方法,另外就是对神经元的权重初始化;而 BP 神经网络训练函数是整个 BP 神经网络形成的引擎,驱动着样本训练过程的执行.
BP 신경망락이경성위응용최위엄범적신경망락모형지일.이인공신경망락시대인뇌진정신경공작적간화생물모형.위료가심대신경망락적리해,이용추도공식래상세분석기최후적수출치화오차.저리지재천술용C어언실현BP신경망락기본모형,재BP신경망락적초시화함수중채용료귀일화처리적방법,령외취시대신경원적권중초시화;이 BP 신경망락훈련함수시정개 BP 신경망락형성적인경,구동착양본훈련과정적집행.
Nowadays BP Neural Network becomes one of the most widely used neural network models, while the artificial neural network is the simplified biological model of human-brain’s real work. For deepening the understanding of neural network, the last output value and error of the neural network is analyzed with the deduction formula. This paper describes basic model of BP neural networks implemented with C language. The normalized method is adopted in the initial function of BP neural network, and the training function of BP neural network is the forming engine of the whole network, and thus it could drive the successful execution of sample training process.