论文以现有的风暴潮灾害研究为基础,通过构建评估指标体系提出了基于粗糙集-支持向量机模型(RS-SVM模型)的风暴潮灾害损失测度方法。RS-SVM模型较好的将二者优点进行结合,消除数据冗余信息的同时可以快速、准确解决非线性问题。在上述研究的基础上,对比单纯使用SVM方法与RSSVM组合使用下的风暴潮经济损失测度效果,发现粗糙集与支持向量机在数据属性约简并进行预测过程中,确保了较高的拟合度与较低的误差率。 Based on the existing storm surge research, the paper proposes a method to measure storm surge disasters based on RS-SVM model by constructing evaluation index system. The RS-SVM model better combines the advantages of both to eliminate the redundant data of the data and solve the nonlinear problem quickly and accurately. On the basis of the above studies, comparing the effect of storm surge economic loss with the combination of SVM and RSSVM, it is found that rough sets and support vector machines can reduce the data attributes and predict the data attributes, Degree and lower error rate.