Agrometeorological Disaster Grading in Guangdong Province Based on Data Mining

Abstract

This study proposes a mining method for meteorological disaster grade rules from the raw data accumulated by meteorological stations using fuzzy association rules. Rules for grading agrometeorological disasters are created and successfully applied to a map. The intention is to mitigate such disasters by understanding their conditions. The procedure described uses the fuzzy c-means clustering algorithm and the Apriori algorithm to mine fuzzy association rules for high-temperature and flooding agrometeorological disasters in Guangdong province, China. In the proposed method, the clustering algorithm does not depend on the membership functions of domain experts. The results show that effective association rules for agrometeorological disasters can be obtained from meteorological data in the long term, even with a lack of prior knowledge. The rules obtained could be used to forecast the grade and region of such disasters in Guangdong province, thus contributing to agrometeorological disaster monitoring and early warning efforts.