首页 | 本学科首页   官方微博 | 高级检索  
     

环境污染的投影寻踪回归预测模型
引用本文:彭荔红,李祚泳,郑文教,肖金树. 环境污染的投影寻踪回归预测模型[J]. 厦门大学学报(自然科学版), 2002, 41(1): 79-83
作者姓名:彭荔红  李祚泳  郑文教  肖金树
作者单位:1. 厦门大学环境科学研究中心,厦门,361005
2. 成都信息工程学院大气电子工程系,成都,610041
3. 厦门大学生物学系,厦门,361005
4. 莆田市环境保护科学研究所,莆田,351100
摘    要:环境污染预测中常用的统计预测法基于“假定-模拟-预测”这样一种证实性数据分析思维模式,难以适应非线性,非正态分布数据分析,投影寻踪技术采用“审视数据-模拟-预测”这样一条探索性数据分析新途径,可以用来解决一定程度上的非线性,非正态问题。应用投影寻踪回归新技术,建立了大气污染物SO2浓度的预测模型,其预测建模样本的拟合合格率为96%,预留预测检验样本预测准确率为80%。高于模糊识别模型的预测精度。

关 键 词:环境污染 投影寻踪回归 环境预测 大气污染 二氧化硫 浓度预测
文章编号:0438-0479(2002)01-0079-05
修稿时间:2001-05-16

Projection Pursuit Regressive Model for Forecasting Environmental Pollution
PENG Li-hong ,LI Zuo-yong ,ZHENG Wen-jiao ,XIAO Jin-shu. Projection Pursuit Regressive Model for Forecasting Environmental Pollution[J]. Journal of Xiamen University(Natural Science), 2002, 41(1): 79-83
Authors:PENG Li-hong   LI Zuo-yong   ZHENG Wen-jiao   XIAO Jin-shu
Affiliation:PENG Li-hong 1,LI Zuo-yong 2,ZHENG Wen-jiao 3,XIAO Jin-shu 4
Abstract:The general statistical method on forecast of environmental pollution, which is based on the analyzing idea of certainty data named as "assumption-simulation-forecast", cannot be applied to nonlinear data and data presenting no normal school. By using the new idea of "examination of data-simulation-forecast", projection pursuit regressive technique can solve the problem of nonlinearity and abnormality to some extent. A projection pursuit regression model on concentration forecast of air pollutant SO 2 was proposed .The results of the forecasts showed that the qualified rates of fitting and forecasting of the model reached 96% and 80%,respectively,both of which were higher than the forecasting precision in fuzzy recognition model.
Keywords:projection pursuit  environmental pollution  forecast  projection pursuit regression
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号