A harmony search-based learning algorithm for epileptic seizure prediction
The learning phase of wavelet neural network entails the task of finding the optimal set of parameter, which includes wavelet activation function, translation centers, dilation parameter, synaptic weight values, and bias terms. Apart from the traditional gradient descent-based approach, metaheuristi...
Saved in:
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Institute of Research and Journals
2016
|
Subjects: | |
Online Access: | http://eprints.uthm.edu.my/5275/1/AJ%202017%20%28747%29.pdf http://eprints.uthm.edu.my/5275/ |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | The learning phase of wavelet neural network entails the task of finding the optimal set of parameter, which includes wavelet activation function, translation centers, dilation parameter, synaptic weight values, and bias terms. Apart from the traditional gradient descent-based approach, metaheuristic algorithms can also be used to determine these parameters. In this work, the harmony search algorithm is employed to find the optimal solution for both synaptic weight values and bias terms in the learning of wavelet neural network. The standard harmony search algorithm is modified accordingly in the aspect of initialization of harmony memory, as well as during the improvisation stage. The proposed harmony search-based learning algorithm is used in the task of epileptic seizure prediction. Simulation results show that the proposed algorithm outperforms other metaheuristic algorithms in terms of sensitivity. |
---|