| 基于多层感知机和RBF转换函数的混合神经网络 |
| |
| 引用本文： | 武妍,王守觉. 基于多层感知机和RBF转换函数的混合神经网络[J]. 计算机工程, 2006, 32(6): 25-27 |
| |
| 作者姓名： | 武妍 王守觉 |
| |
| 作者单位： | 同济大学计算机科学与工程系,上海,200092;同济大学半导体与信息技术研究所,上海,200092;同济大学半导体与信息技术研究所,上海,200092;中国科学院半导体研究所神经网络实验室,北京,100083 |
| |
| 基金项目： | 中国科学院资助项目;上海市博士后科研项目 |
| |
| 摘 要： | 为了更有效地优化前向神经网络的求解能力，提出了一种新的综合的转换函数，将多层感知机和RBF神经网络更有机地结合起来，以产生灵活的决策边界。在此基础上推导出了相应的学习算法。并通过对实际的模式分类问题的仿真，将文中的方法与带动量项BP算法、CSFN、RBF等算法进行了比较，验证了其有效性。
|
| 关 键 词： | 转换函数 径向基函数 多层感知机 混合网络 学习算法 |
| 文章编号： | 1000-3428（2006）06-0025-03 |
| 收稿时间： | 2005-03-17 |
| 修稿时间： | 2005-03-17 |
Hybrid Neural Network Based on Transfer Functions of Multilayer Perception and Radial Basis Function |
| |
| WU Yan,WANG Shoujue. Hybrid Neural Network Based on Transfer Functions of Multilayer Perception and Radial Basis Function[J]. Computer Engineering, 2006, 32(6): 25-27 |
| |
| Authors: | WU Yan WANG Shoujue |
| |
| Affiliation: | 1. Dept. of Computer Science and Engineering, Tongji University, Shanghai 200092; 2. Institute of Semiconductors and Information Technology Tongji University, Shanghai 200092; 3. Lab of Artificial Neural Networks, Institute of Semiconductors, CAS, Beijing 100083 |
| |
| Abstract: | In order to effectively optimizing the solution of feed-torward neural network, a new general transfer function is proposed that effectively unifies the inputs of multilayer perception and radial basis function to provide flexible decision border. A new algorithm based on gradient descent and error propagation is proposed. Several pattern classification example simulations are made to verify the validity of the proposed algorithm by comparing the proposed transfer function and learning algorithm with BP algorithm adding momentum term, CSFN and RBF. |
| |
| Keywords: | Transfer fi, nction, Radial basis function Multilayer perception Hybrid network Learning algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
