中国是一个受震灾影响严重的国家，中国的地震具有高频、高强度、浅源三大特点。其地理特征又决定了东西部所受震灾存在差异，东部频率相对低，强度相对大，西部恰恰相反。再加上东西部经济人口也存在显著差异，地震灾害带来的后果复杂多变，在政策的制定上需要因地制宜。本文主要处理两个方面的问题。一方面建立多元线性回归模型，旨在利用地震初期能够立刻获取或者估算到的数据，估计出经济损失，为政府应对震灾提供参考。另一方面利用聚类分析，根据多个变量将地震进行分类，分别分析每个类型的地震，为政府制定不同类型应急计划提供参考。

China is a country which is seriously affected by the earthquake. China's earthquake has three prominent characteristics which are high frequency, high intensity and small depth of the earthquakes. Its geographical characteristics also determine the difference of the earthquake between the east and west. In the east, frequency is relatively low, and the intensity is relatively large, while the west is the opposite. Coupled with the significant differences in the economic population of the East and the West, the consequences of earthquake disasters are complex and changeable, and the formulation of policies needs to be tailored to local conditions.

This paper mainly deals with two aspects. On the one hand, a multilinear regression model is established, which aims to estimate the economic loss and provide reference for the government to deal with the earthquake by using the data that can be obtained or estimated immediately at the beginning of the earthquake. On the other hand, clustering analysis is used to classify earthquakes according to multiple variables, analyzing each type of earthquake separately, and providing reference for the government to formulate different types of emergency plans.