Skip Nav Destination
ASME Press Select Proceedings
Intelligent Engineering Systems through Artificial Neural Networks, Volume 16
ISBN-10:
0791802566
No. of Pages:
1000
Publisher:
ASME Press
Publication date:
2006
eBook Chapter
91 A New Exploratory Neural Network Training Method
By
Shamsuddin Ahmed
Kazakhstan Institute of Management Economics and Strategic Research , Almaty .Edith Cowan University , SOEM, 100-Joondalup drive, Perth, WA-6027 , Australia .
,
Shamsuddin Ahmed
Search for other works by this author on:
Jim Cross
Edith Cowan University , SOEM, 100-Joondalup drive, Perth, WA-6027 , Australia .
,
Jim Cross
Search for other works by this author on:
Salim Bouzerdoum
University of Wollongong , SECTE, Wollongong, NSW-2522 , Australia
Salim Bouzerdoum
Search for other works by this author on:
Page Count:
6
-
Published:2006
Citation
Ahmed, S, Cross, J, & Bouzerdoum, S. "A New Exploratory Neural Network Training Method." Intelligent Engineering Systems through Artificial Neural Networks, Volume 16. Ed. Dagli, CH, Buczak, AL, Enke, DL, Embrechts, M, & Ersoy, O. ASME Press, 2006.
Download citation file:
A new exploratory self-adaptive derivative free training algorithm is developed. It only evaluates error function that is reduced to a set of sub-problems in a constrained search space and the search directions follow rectilinear moves. To accelerate the training algorithm, an interpolation search is developed that determines the best learning rates. The constrained interpolation search decides the best learning rates such that the direction of search is not deceived in locating the minimum trajectory of the error function.
The proposed algorithm is practical when the error function is ill conditioned implying that the Hessian matrix property is unstable, or the...
Abstract
1 Exploratory Training
2 Exploratory Training Algorithm
3 Convergence of the Training Method
4 Analysis with the XOR Problem
5 Training Results
6 Discussions
References
This content is only available via PDF.
You do not currently have access to this chapter.
Email alerts
Related Chapters
An Interpolation FFT Algorithm for Harmonics Analysis Based on Maximum Side Lobe Decay Cosine Windows
International Conference on Mechanical and Electrical Technology, 3rd, (ICMET-China 2011), Volumes 1–3
Discovering Similar Segments by Time-Parameterized Trajectory Clustering
International Conference on Mechanical and Electrical Technology, 3rd, (ICMET-China 2011), Volumes 1–3
Manipulability-Maximizing SMP Scheme
Robot Manipulator Redundancy Resolution
Research of Improved BP Neural Network in Intrusion Detection
Proceedings of the International Conference on Internet Technology and Security
Related Articles
Solving Nonlinear Fractional Integro-Differential Equations of Volterra Type Using Novel Mathematical Matrices
J. Comput. Nonlinear Dynam (November,2015)
The Discrete Adjoint Gradient Computation for Optimization Problems in Multibody Dynamics
J. Comput. Nonlinear Dynam (May,2017)
Mechanism Design with MP-Neural Networks
J. Mech. Des (December,1998)