A class of unconstrained minimization methods for neural network training*
作者:
L. Grippo,
期刊:
Optimization Methods and Software
(Taylor Available online 1994)
卷期:
Volume 4,
issue 2
页码: 135-150
ISSN:1055-6788
年代: 1994
DOI:10.1080/10556789408805583
出版商: Gordon and Breach Science Publishers
关键词: Neural networks, training algorithms, unconstrained minimization
数据来源: Taylor
摘要:
In this paper the problem of neural network training is formulated as the unconstrained minimization of a sum of differentiate error terms on the output space. For problems of this form we consider solution algorithms of the backpropagation-type, where the gradient evaluation is split into different steps, and we state sufficient convergence conditions that exploit the special structure of the objective function. Then we define a globally convergent algorithm that uses the knowledge of the overall error function for the computation of the learning rates. Potential advantages and possible shortcomings of this approach, in comparison with alternative approaches are discussed.
点击下载:
PDF (558KB)
返 回