| 基于神经网络的非线性PCA方法 |
| |
| 引用本文： | 孔薇,杨杰. 基于神经网络的非线性PCA方法[J]. 计算机仿真, 2003, 20(7): 65-67,96 |
| |
| 作者姓名： | 孔薇 杨杰 |
| |
| 作者单位： | 上海交通大学图像处理与模式识别研究所,上海,200030 |
| |
| 摘 要： | 该文采用基于正交最小二乘方法(OLS)的径向基函数(RBF)神经网络进行非线性主元分析(NLPCA)算法的训练，提高了训练速度，且不存在局部最优问题。将其应用到聚丙烯生产的高维非线性数据相关特性的提取中，仿真试验显示这种NLPCA方法提高了熔融指数(MI)的预报精度，具有实际应用价值。
|
| 关 键 词： | 非线性主元分析方法 神经网络 PCA 径向基函数 正交最小二乘方法 |
| 文章编号： | 1006-9348(2003)07-0065-03 |
Applications of Nonlinear PCA Based on Neural Network in Prediction of Melt Index |
| |
| KONG Wei,YANG Jie. Applications of Nonlinear PCA Based on Neural Network in Prediction of Melt Index[J]. Computer Simulation, 2003, 20(7): 65-67,96 |
| |
| Authors: | KONG Wei YANG Jie |
| |
| Abstract: | An approach to nonlinear principal component analysis (NLPCA) which using RBF neural network based on OLS algorithm has been developed for the reason that linear PCA can't extract nonlinear features, and this algorithm improves the training speed without local optimization. It is used to extract the correlations among the high-dimension nonlinear data of polypropylene manufacture. The simulation results show that NLPCA successfully reduces the dimensions and effectively improves the precision of the prediction of melt index (MI). |
| |
| Keywords: | Principal component analysis (PCA) Nonlinear Principal component analysis (NLPCA) RBF Neural Network Orthogonal Least Squares (OLS) Melt Index (MI) |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
