基于PCA的聚类分析在汶川地震灾情分类中的应用

Application of Cluster Analysis Based on Principal Component Analysis to Classification of Disaster in Wenchuan Earthquake

利用主成分分析(PCA)运用SPSS软件对汶川地震36个严重受灾县市的8个灾情指标进行了综合分析,得到了累积贡献率为83.403%的3个主成分及其得分;然后,基于3个主成分的得分采用聚类分析对汶川地震36个严重受灾县市进行了分类;得到了全面、合理和科学的分类结果。

The comprehensive analysis is conducted on 8 disaster indicators of 36 seriously disaster-stricken counties and Cities in Wenchuan Earthquake by principal component analysis and by SPSS software, three principal components and scores of cumulative contribution rate of 83. 403% are obtained. Then,the classification for 36 seriously disaster-stricken counties and cities in Wenchuan Earthquake is conducted by cluster analysis based on the scores of three principal components. The comprehensive, rational and scientific classification results are obtained.