基于小波神经网络的Shibor预测研究

A Prediction Research on Shibor based on Wavelet Neural Network

上海银行间同业拆放利率(Shibor)的推出是中国利率市场化重要的一步。在阐述了Shibor的背景、功能以及对经济发展的重大意义之后,分别建立了小波神经网络和回归时间序列组合模型对2周品种Shibor进行预测对比分析,研究结果表明,小波神经网络的拟合和预测精度较高,具有一定的科学性和实用性。

Pushing out of Shanghai inter bank offered rate （Shibor） is the important step for China＇ s interest rate marketization. This paper describes the background and functions of Shibor and the important meaning for economic development. The author established respectively wavelet neural network and retrogression time sequence combination model to the two-week variety of Shibor for prediction comparative analysis and found that the wavelet neural network is more accurate and has been scientific and practical.