A Pruning Algorithm Based on Relevancy Index of Hidden Neurons Outputs
Journal Title: Engineering, Technology & Applied Science Research - Year 2016, Vol 6, Issue 4
Abstract
Choosing the training algorithm and determining the architecture of artificial neural networks are very important issues with large application. There are no general methods which permit the estimation of the adequate neural networks size. In order to achieve this goal, a pruning algorithm based on the relevancy index of hidden neurons outputs is developed in this paper. The relevancy index depends on the output amplitude of each hidden neuron and estimates his contribution on the learning process. This method is validated with an academic example and it is tested on a wind turbine modeling problem. Compared with two modified versions of Optimal Brain Surgeon (OBS) algorithm, the developed approach gives interesting results.
Authors and Affiliations
S. Abid, M. Chtourou, M. Djemel
Keywords
Related Articles
- EP ID EP109679
- DOI -
- Views 260
- Downloads 0
A Novel Structure for Vector Control of Symmetrical Six-Phase Induction Machines with Three Current Sensors
Vector control is one of the most popular multiphase induction machine drive methods due to its dynamic performance. Since all phase currents are needed in vector control drive, by increasing phase numbers, the number of...
Feature Set Fusion for Spoof Iris Detection
Iris recognition is considered as one of the most promising noninvasive biometric systems providing automated human identification. Numerous programs, like unique ID program in India - Aadhar, include iris biometric to p...
A Comparison Between Inter-Asterisk eXchange Protocol and Jingle Protocol: Session Time
Over the last few years, many multimedia conferencing and Voice over Internet Protocol (VoIP) applications have been developed due to the use of signaling protocols in providing video, audio and text chatting services be...
Studies of the Properties and Microstructure of Heat Treated 0.27% C and 0.84% Mn Steel
In the present work, different heat treatments like hardening with different cooling rates followed by tempering at different temperatures have been performed. The material used in this study is carbon steel of 0.27% C a...
Variable Step Size Maximum Correntropy Criteria Based Adaptive Filtering Algorithm
Maximum correntropy criterion (MCC) based adaptive filters are found to be robust against impulsive interference. This paper proposes a novel MCC based adaptive filter with variable step size in order to obtain improved...