论文部分内容阅读
利用主成分分析和聚类分析,介绍了东亚飞蝗发生程度及对不同发生程度进行分类的方法。主成分分析选出秋残蝗最高密度、夏蝗达标面积、秋残蝗面积、夏蝗发生面积等4个主因子。这4个因子能在84.56%的程度上反映东亚飞蝗的发生程度。聚类分析将东亚飞蝗20年的资料,在卡方距离近1.68处聚成3大类,分析结果与实际情况非常吻合。依据聚类分析结果可对东亚飞蝗进行分类综合防治。
Using principal component analysis and cluster analysis, introduced the degree of occurrence of oriental migratory locusts and the classification of different occurrences. Principal component analysis selected the highest density of autumn residue locusts, summer locust standard area, autumn residue locust area, summer locust occurrence area and other four main factors. These four factors can reflect the degree of occurrence of oriental migratory locusts to the extent of 84.56%. Cluster analysis of East Asian migratory locust 20 years of data, in the chi square distance 1.68 gathered into three categories, the analysis results and the actual situation is very consistent. Based on the results of cluster analysis, the migratory locusts of East Asia can be classified and comprehensively controlled.