摘要
该文提出一种基于构造型神经网络的最大密度覆盖分类算法,该算法直接从样本数据本身入手,通过引入一个密度估计函数对样本数据进行聚类分析,找出同类样本中具有最大密度的样本数据点,然后在特征空间里作超平面与球面相交,得到一个球面领域覆盖,从而将神经网络训练问题转化为点集覆盖问题.该算法有效地克服了传统神经网络训练时间长、学习复杂的问题,同时也考虑了神经网络规模的优化问题,实验证明了该算法的有效性.
A new maximum density covering classitication algorithm based on constructive neural networks is proposed in this paper, which starts with the sample data directly and clustering analysis is proceeded with to find a sample with the maximum density and then the intersection between the positive half-space of the hyperplane and sphere, called “sphere neighborhood”, is obtained, by which the training problem of neural networks can be transformed into the covering problem of a point set. Thus the new algorithm can reduce the long training time and learning complexity of traditional neural networks. At the same time, the optimization of the neural network is also considered and computer simulation results show that the proposed neural network is quite efficient.
出处
《计算机学报》
EI
CSCD
北大核心
2005年第9期1519-1523,共5页
Chinese Journal of Computers
基金
国家"八六三"高技术研究发展计划项目基金(2001AA413130)资助
关键词
模式识别
神经网络
最大密度覆盖
M—P神经元
构造型神经网络
pattern recognition
neural networks
maximum density covering
M-P neuron
constructive neural networks
