论文信息 - A normalised complex backpropagation algorithm

A normalised complex backpropagation algorithm

A backpropagation based algorithm for training nonlinear complex valued feed-forward neural networks employed as nonlinear adaptive filters is derived. The proposed normalised complex backpropagation (NCBP) algorithm is an improvement on the complex backpropagation (CBP) algorithm by including an adaptive normalised learning rate. This is achieved by performing a minisation of the complex-valued instantaneous output error that has been expanded via a Taylor series expansion. The proposed algorithm is applicable to any complex-valued nonlinear architecture. Experiments on complex coloured and nonlinear signals confirm that the NCBP algorithm outperforms the standard CBP algorithm.

Danilo P. Mandic | Andrew I. Hanna

[1] B. Widrow,et al. The complex LMS algorithm , 1975, Proceedings of the IEEE.

[2] Danilo P. Mandic,et al. Recurrent Neural Networks for Prediction: Learning Algorithms, Architectures and Stability , 2001 .

[3] Cris Koutsougeras,et al. Complex domain backpropagation , 1992 .

[4] Danilo P. Mandic,et al. A normalised backpropagation learning algorithm for multilayer feed-forward neural adaptive filters , 2001, Neural Networks for Signal Processing XI: Proceedings of the 2001 IEEE Signal Processing Society Workshop (IEEE Cat. No.01TH8584).

[5] Danilo P. Mandic,et al. Recurrent Neural Networks for Prediction: Learning Algorithms, Architectures and Stability , 2001 .

[6] Tülay Adali,et al. Fully Complex Multi-Layer Perceptron Network for Nonlinear Signal Processing , 2002, J. VLSI Signal Process..

[7] Francesco Piazza,et al. On the complex backpropagation algorithm , 1992, IEEE Trans. Signal Process..