A Proposed Image Structure of Multiwavelet Network | ||
| Journal of Engineering and Sustainable Development | ||
| Article 1, Volume 18, Issue 6, December 2014, Pages 233-245 | ||
| Authors | ||
| Wael Hussein Zayer; Ahmed Thamer Radi | ||
| Abstract | ||
| The combination of wavelet theory and neural networks has lead to the development of wavelet networks. Wavelet networks are feed-forward neural networks using wavelets as activation function. Wavelet networks have been used in classification and identification problems with some success. The strength of wavelet networks lies in their capabilities of catching essential features in ''frequency-rich" signals. In wavelet networks, both the position and the dilation of the wavelets are optimized besides the weights. Proposed multi wavelet network are used in identification problems of nonlinear systems. A multiwavelet network is constructed as an alternative to a neural network to approximate a nonlinear system. | ||
| Keywords | ||
| wavelet transform; Multiwavelet Network; Neural network; Wavelet networks | ||
|
Statistics Article View: 105 PDF Download: 42 |
||