期刊文献+

重置变结构神经网络的算法 被引量:1

Algorithm of revisable structural neural network
下载PDF
导出
摘要 神经网络的结构直接影响到网络性能的优劣,进而影响其推广使用,是目前研究的焦点之一.尝试将重置算法应用于经典BP神经网络的结构优化,研究了重置算法中最佳重置时间的性质,同时提出一种重置变结构经典BP神经网络. The structure of neural network, as a hotspot of research in artificial neural network(ANN), will affect its performance and application directly. This article applies the early-restart algorithm to the optimization of neural network's structure, studies the character of optimal restart time in the algorithm, and advances the revisable structure of classic BP neural network based on early-restart algorithm.
出处 《山东理工大学学报(自然科学版)》 CAS 2004年第2期25-28,共4页 Journal of Shandong University of Technology:Natural Science Edition
关键词 重置变结构 神经网络 算法 非线性动态系统 自学习 revisable structure early-restart algorithm neural network
  • 相关文献

参考文献16

  • 1Rumelhart D E,McClelland J L. Parallel and distributed processing(PDF): Exploration in themicrostructure of the cognition[M]. MIT press, Cambridge, MA, 1986.
  • 2Gorman R P,Sejnowski T J. Learned classification of sonar target using a massively paralled network[J]. IEEE Trans. Account, Speech signal processing, 1988, 36:1 135-1 140.
  • 3Poth M V. Neural networks for extraction of weak targets in high clutter environments[ J ]. IEEE Trans. System, Man, Cyber, 1989, 19:1 210-1 217.
  • 4Narendra K S, Parthasarathy K. Identification and control of dynamical systems using neural networks[J]. IEEE Trans. Neural Networks,1999, 1:27-43.
  • 5Nikoonahad M, Liu D C. Medical ultrasound imaging using neural networks[J]. Electron, 1990, 26: 545-546.
  • 6郑立群,吴育华,周伯康,夏庆.人工神经网络方法在投资风险评价中的应用[J].管理科学学报,1999,2(4):93-95. 被引量:16
  • 7张新红.基于神经网络的高技术项目投资风险综合评价模型[J].情报理论与实践,2001,24(5):377-379. 被引量:7
  • 8何述东,瞿坦,黄献青,黄心汉.多层前向神经网络结构的研究进展[J].控制理论与应用,1998,15(3):313-319. 被引量:37
  • 9Saratchandran P. Dynamic programming approach to optimal weights selection in multiplayer neural networks[J]. IEEE Trans. Neural Network, 1991,2: 465-467.
  • 10Frean M. The upstart algorithm: A method for constructing and training feed forward neural networks[J ]. Neural Computation. 1990, 2(2): 198-209.

二级参考文献20

共引文献56

同被引文献9

引证文献1

二级引证文献2

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部