Abstract: Natural disasters injure people and damage property. Because the weather is related to many factors, it is difficult to forecast accurately the disastrous weather. Based on multiclassifiers, a cooperative data mining method is proposed. A cooperative classification model is designed and implemented, which is composed of base classifiers and an integration classifier. The model is used to mine local area meteorological data. Experimental results show that the model has high classification accuracy and efficient ability.

Key words: data mining; K-Nearest Neighbor algorithm; cooperative; multiclassifier model; local area meteorological data