Multiscale BiLinear Recurrent Neural Network with an Adaptive Learning Algorithm
In this paper, a wavelet-based neural network architecture called the Multiscale BiLinear Recurrent Neural Network with an adaptive learning algorithm (M-BLRNN(AL)) is proposed. The proposed M-BLRNN(AL) is formulated by a combination of several BiLinear Recurrent Neural Network (BLRNN) models in which each model is employed for predicting the signal at a certain level obtained by a wavelet transform. The learning process is further improved by applying an adaptive learning algorithm at each resolution level. The proposed M-BLRNN(AL) is applied to the long-term prediction of MPEG VBR video traffic data. Experiments and results on several MPEG data sets show that the proposed M-BLRNN(AL) outperforms the traditional MultiLayer Perceptron Type Neural Network (MLPNN), the BLRNN, and the original M-BLRNN in terms of the normalized mean square error (NMSE).
KeywordsRecurrent Neural Network Resolution Level Adaptive Learning Asynchronous Transfer Mode Normalize Mean Square Error
Unable to display preview. Download preview PDF.
- 3.Bhattacharya, A., Parlos, A.G., Atiya, A.F.: Prediction of MPEG-coded Video Source Traffic using Recurrent Neural Networks. IEEE Trans. on Acoustics, Speech, and Signal Processing 51, 2177–2190 (2003)Google Scholar
- 4.Park, D.C., Zhu, Y.: Bilinear Recurrent Neural Network. IEEE ICNN 3, 1459–1464 (1994)Google Scholar
- 6.Shensa, M.J.: The Discrete Wavelet Transform: Wedding the À Trous and Mallat Algorithms. IEEE Trans. Signal Proc. 10, 2463–2482 (1992)Google Scholar
- 10.Parlos, A.G., Rais, O.T., Atiya, A.F.: Multi-step-ahead Prediction using Dynamic Recurrent Neural Networks. In: IJCNN 1999. Int. Joint Conf. on Neural Networks, vol. 1, pp. 349–352 (1999)Google Scholar