期刊文献+

改进的RBF神经网络模式分类方法应用研究 被引量:6

Applications of Pattern Classification Methods Based on Improved RBF Neural Networks
下载PDF
导出
摘要 经典的 Bayes分类方法一般需要事先对样本的分布特性作出假设 ,当假设模型与样本实际分布情况不相符时 ,就难以得到较高的分类精度。当处理同类别多区域样本分布问题 ,例如变标签问题时 ,距离判别、Fisher判别、k-近邻分类、分段线性分类等统计分析方法遇到困难。双螺旋问题不仅使统计方法受到挑战 ,更使人们对一般前向多层神经网络的能力提出疑问。本文提出了改进的 RBF神经网络结构、核函数个数、位置与宽度优化算法。该算法的计算复杂性与一般前向三层LBF网络所用的误差反传算法大致相同。核函数生成既考虑了训练集样本自身的类别因素 ,又考虑了错分样本与邻近类别的关系。一个核函数的最终保留与否根据其对提高测试集分类正确率的贡献大小来决定。同时实验验证了两层 LBF网络对提高改进的 RBF网络分类正确率的极端重要性。大量应用实例表明 ,与前向三层 RBF网络和前向三层 LBF网络相比 ,该 IRBF网络具有收敛速度快、分类精度高、易于得到最小结构、在学习过程中不易陷入局部极小点等优点 。 Generally speaking, the classical Bayes classification methods must hypothesize what distribution a random variable be subject to before analyzing. It is impossible to get a high correct rate if the selected model is not agreement with the true one. The statistic approaches, for example, distance, Fisher, k nearest neighbor, wise linear classifiers, fail to solve multi regional distributions such as the alternate table problems. Not only does the two spiral problem give a challenge to the stati...
出处 《华东理工大学学报(自然科学版)》 CAS CSCD 北大核心 2001年第6期684-692,共9页 Journal of East China University of Science and Technology
基金 生物反应器国家重点实验室开放课题 ( SK0 0 -0 7)
关键词 径基函数 线性基本函数 神经网络 模式分类 网络结构 核函数 分类方法 学习算法 radial basis function linear basis function neural networks pattern classification
  • 引文网络
  • 相关文献

参考文献1

二级参考文献17

  • 1[1]Gao Daqi, Wu Shouyi. An optimization method for the topolog ical structures of feed-forward multi-layer neural networks[J]. Pattern Reco gnition, 1998,31(9):1 337-1 342.
  • 2[2]Poggio T, Girosi F. Networks for approximation and learning[J]. P roceedings of IEEE, 1990,78(9):1 481-1 495.
  • 3[3]Lippmann R P. Pattern classification using neural networks[J]. IE EE Communication Magazine, 1989,11:47-64.
  • 4[4]Leonardis A, Bischof H. An efficient MDL-based construction of RBF networks[J]. Neural Networks, 1998,11(4):963-973.
  • 5[5]Musavi M T, Ahmed W, Chan K H, et al. On the training of radial ba sis function classifiers[J]. Neural Networks, 1992,5(3):595-603.
  • 6[6]Bishop C M. Neural Networks for Pattern Recognition[M]. Oxford: C larendon Press, 1998.
  • 7[7]Chen S, Cowan C F N, Grant P M. Orthogonal least squares learning a lgorithm for radial basis function networks[J]. IEEE Transactions on Neural Ne tworks, 1991,2(2):302-309.
  • 8[8]Roy A, Govil S, Miranda P. An algorithm to generate radial basis fu nction(RBF)-like nets for classification problems[J]. Neural Networks, 1995,[ STHZ]8(2):179-201.
  • 9[9]Karayiannis N B. Reformulated radial basis neural networks trained by gradient descent[J]. IEEE Transactions on Neural Networks, 1999,10(3):657-671.
  • 10[10]Karayiannis N B, Mi G W. Growing radial basis neural networks: merging s upervised and unsupervised learning with network growth techniques[J]. IEEE Tr ansactions on Neural Networks, 1997,8(6):1 492-1 506.

共引文献6

同被引文献52

引证文献6

二级引证文献25

;
使用帮助 返回顶部