摘要
本文主要应用了“对分移动窗口法”和“特征的逐次筛选法”。所得分类结果基本上与实际地震活动情况一致,后验预测检验也符合实际。在使用筛选后的特征分类时发现聚类方法对其所使用的特征有某种程度的依赖性。
The webster removable window method, modified graphic theory and 'the progressive feature screening' method are used in the paper. The results obtained are bascally consistent with the actual seismic activities. The clustering methods have a certain dependence on features when using the classified features after screening.
出处
《中国地震》
CSCD
北大核心
1989年第4期6-12,共7页
Earthquake Research in China
基金
地震学科学联合基金
