论文信息 - Modeling Dynamical Systems by Error Correction Neural Networks

Modeling Dynamical Systems by Error Correction Neural Networks

We introduce a new time delay recurrent neural network called ECNN, which includes the last model error as an additional input. Hence, the learning can interpret the models misspecification as an external shock which can be used to guide the model dynamics afterwards.

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

[2] James L. McClelland,et al. Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations , 1986 .

[3] Barak A. Pearlmutter. Gradient calculations for dynamic recurrent neural networks: a survey , 1995, IEEE Trans. Neural Networks.

[4] John F. Kolen,et al. Field Guide to Dynamical Recurrent Networks , 2001 .

[5] Paul J. Werbos,et al. The Roots of Backpropagation: From Ordered Derivatives to Neural Networks and Political Forecasting , 1994 .

[6] Paul J. Werbos,et al. The roots of backpropagation , 1994 .

[7] J. D. Farmer,et al. State space reconstruction in the presence of noise" Physica D , 1991 .

[8] Jeffrey L. Elman,et al. Finding Structure in Time , 1990, Cogn. Sci..

[9] Simon Haykin,et al. Neural Networks: A Comprehensive Foundation , 1998 .

[10] Geoffrey E. Hinton,et al. Learning internal representations by error propagation , 1986 .

[11] T. Lai. Time series analysis univariate and multivariate methods , 1991 .

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

[13] Martin Casdagli,et al. An analytic approach to practical state space reconstruction , 1992 .

[14] William W. S. Wei,et al. Time series analysis - univariate and multivariate methods , 1989 .

[15] Lakhmi C. Jain,et al. Recurrent Neural Networks: Design and Applications , 1999 .