Elman神经网络在银行间同业拆借利率分析中的应用

The use of the Elman Neural Network in the Inter-Bank Borrowing and Lending Interest Rate

利用Elman网络建立了同业拆借利率的神经网络模型,根据msereg性能函数确定了神经网络的输入层个数,采用不同的算法训练网络,从中选出L-M算法是最为快速和准确的。经所提算法训练过的网络在处理复杂时间序列方面具有很好的学习能力和泛化能力。

The paper constructs a neural network of the inter-bank borrowing and lengding interest rate using Elman network.Firstly,it uses msereg utility function to count the number of the input layer,then makes use of different algorithms to train the network,and finds that L-M method is the fastest and the most correct.The demonstration reveals that the method the paper suggests can have good performance in study and is very useful.