摘要
利用神经网络和粗糙集处理不确定性问题的优势,提出一种粗糙集结合神经网络进行森林火灾预测模型。通过与传统预测模型相比较,证明了该方法的有效性。
A new method for forest fire alarm based on the neural network and rough set is proposed. Rough-set is used to preprocess the data by simplification and reduction, which can minimize the scale of attribute by removing the unnecessary attributes, so that enhance the convergence speed and accuracy of neural network. The efficiency of this method is proved by comparing of performance with the traditional method.
出处
《武汉大学学报（信息科学版）》
EI
CSCD
北大核心
2006年第8期720-723,共4页
Geomatics and Information Science of Wuhan University
基金
湖北省自然科学基金资助项目(2001ABB041)
