To create a new prediction model, the unbiased GM （1,1） model is optimized by the five-point slide method in this paper. Then, based on the occurrence areas of dce blast in Enshi District during 1995 -2004, the new ...To create a new prediction model, the unbiased GM （1,1） model is optimized by the five-point slide method in this paper. Then, based on the occurrence areas of dce blast in Enshi District during 1995 -2004, the new model and unbiased GM （1, 1 ） model are applied to predict the occurrence areas of rice blast during 2005 -2010. Predicting outcomes show that the prediction accuracy of five-point unbiased sliding optimized GM （1, 1 ） model is higher than the unbiased GM （1,1） model. Finally, combined with the prediction results, the author provides some suggestion for Enshi District in the prevention and control of rice blast in 2010.展开更多
针对区间值数据的数据聚类问题,根据可拓学关联函数的定义,提出可拓距离的概念来度量数据之间的距离,利用K近邻的思想,根据可拓距离的大小对数据集的目标属性进行投票选择进行分类,设计了可拓K近邻算法(Extension K Nearest Neighbor,EK...针对区间值数据的数据聚类问题,根据可拓学关联函数的定义,提出可拓距离的概念来度量数据之间的距离,利用K近邻的思想,根据可拓距离的大小对数据集的目标属性进行投票选择进行分类,设计了可拓K近邻算法(Extension K Nearest Neighbor,EKNN)。最后利用UCI的两个基准数据集Iris植物样本数据和糖尿病数据库PIDD进行验证,首先通过免疫网络约简算法对条件属性进行最小属性约简,然后利用EKNN算法分析和比较不同最小约简属性下的分类准确率。展开更多
基金Supported by Science Research Project of Department of Education of Hubei Province (B20092901)~~
文摘To create a new prediction model, the unbiased GM （1,1） model is optimized by the five-point slide method in this paper. Then, based on the occurrence areas of dce blast in Enshi District during 1995 -2004, the new model and unbiased GM （1, 1 ） model are applied to predict the occurrence areas of rice blast during 2005 -2010. Predicting outcomes show that the prediction accuracy of five-point unbiased sliding optimized GM （1, 1 ） model is higher than the unbiased GM （1,1） model. Finally, combined with the prediction results, the author provides some suggestion for Enshi District in the prevention and control of rice blast in 2010.
文摘针对区间值数据的数据聚类问题,根据可拓学关联函数的定义,提出可拓距离的概念来度量数据之间的距离,利用K近邻的思想,根据可拓距离的大小对数据集的目标属性进行投票选择进行分类,设计了可拓K近邻算法(Extension K Nearest Neighbor,EKNN)。最后利用UCI的两个基准数据集Iris植物样本数据和糖尿病数据库PIDD进行验证,首先通过免疫网络约简算法对条件属性进行最小属性约简,然后利用EKNN算法分析和比较不同最小约简属性下的分类准确率。
