论文信息 - RBF Network-Based Chaotic Time Series Prediction and It's Application in Foreign Exchange Market

RBF Network-Based Chaotic Time Series Prediction and It's Application in Foreign Exchange Market

The foreign exchange market is a chaotic dynamic system. We apply the RBF network-based chaotic time series prediction on the daily USD/RMB exchange rate. We apply the RBF network and phase space reconstruction to find the optimal embedding dimension in the foreign exchange market from the point view of forecasting. We find that the optimal embedding dimension is 10. As a result the dimension of the attractor of the market is about in the interval between 4 and 5. Finally, we use the optimal embedding dimension to implement the prediction.

Li-li Ma | Xu-song Xu | Xueqing Xu | Lilian Ma

[1] Thanasis Stengos,et al. Some evidence concerning macroeconomic chaos , 1988 .

[2] Xu Xu-song. The Fractal Dimension of Shanghai Stock Market Based on RBF Neural Networks , 2003 .

[3] Adam Krzyzak,et al. Nonparametric estimation and classification using radial basis function nets and empirical risk minimization , 1996, IEEE Trans. Neural Networks.

[4] H. Kantz,et al. Nonlinear time series analysis , 1997 .

[5] B. LeBaron,et al. Nonlinear Dynamics and Stock Returns , 2021, Cycles and Chaos in Economic Equilibrium.

[6] James P. Crutchfield,et al. Geometry from a Time Series , 1980 .

[7] F. Takens. Detecting strange attractors in turbulence , 1981 .

[8] Ye Dong. Application of Radial Basis Function Neural Network to Pattern Classification of Stock Changing Tendency , 1999 .

[9] Martin Burger,et al. Error Bounds for Approximation with Neural Networks , 2001, J. Approx. Theory.