Abstract: In the paper new hybrid cascade wavelet-neural network and its learning algorithm in batch and on-line mode are proposed. Such architecture can be used for solving prediction and emulation non-stationary non-linear time series under current and a-priori uncertenity. The computational experiments confirm the effectiveness of developed approach.

Keywords: cascade neural network, wavelet-neuron, learning algorithm, prediction, emulation.

ACM Classification Keywords: I.2.6 Learning – Connectionism and neural nets.

Link:

HYBRID CASCADE NEURAL NETWORK BASED ON WAVELET-NEURON

Yevgeniy Bodyanskiy, Oleksandra Kharchenko, Olena Vynokurova

http://foibg.com/ijita/vol18/ijita18-4-p03.pdf