基于MDP-ISAP的ENN分类算法及其应用  

MDP-ISAP with ENN-based Classification Algorithm and Its Application

作　　者：夏崇坤 苏成利[1] 曹江涛[1] 李平[1] 

机构地区：[1]辽宁石油化工大学信息与控制工程学院,抚顺113001

出　　处：《控制工程》2017年第9期1958-1964,共7页Control Engineering of China

基　　金：国家自然科学基金(61203021);辽宁省科技攻关项目(2011216011);辽宁省自然科学基金项目(2013020024);辽宁省高等学校杰出青年学者成长计划(LJQ2015061)

摘　　要：针对可拓神经网络(ENN)对复杂样本数据分类效果较差的问题,提出一种融合边界判别投影(MDP)和改进半监督近邻传播(ISAP)的新型ENN分类算法。首先,使用边界判别投影对原始数据进行降维,提取关键特征。其次,在低维特征空间进行聚类分析,利用近邻传播聚类筛选出有效训练样本,并通过ISAP聚类寻优获得样本类中心,作为初始类中心,在此基础上构建新的分类器。最后,将其应用于复杂化工过程中高密度聚乙烯(HDPE)的熔融指数预测,取得了较好效果。To solve the classification problem of complicated sample data, a novel extension neural network （ENN） classification algorithm based on margin discriminant projection （MDP） and improved semi-supervised affinity propagation （ISAP） is proposed. At first, the MDP method is used to reduce the dimensions of initial data and extract key features. Secondly, clustering analysis is conducted in the dimension-reduced feature space. On one hand, the affinity propagation （AP） clustering method is used to select sufficient training samples; on the other hand, the ISAP clustering method is used to obtain class centers which are used as the initial cluster centers, and then an improved ENN classifier is constructed. Finally, the proposed method is employed to the high density polyethylene （HDPE） in the complicated chemical process. The results also show that, compared with the ENN, the proposed algorithm has better performance.

关 键 词：可拓神经网络 边界判别投影 降维 近邻传播算法 HDPE过程 预测 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]