The recognition of the stored-grain pests is a multi-feature and multi-compound degree classification of various pests. It is very important to determine the objective weights of features automatically in the classification of the stored grain pests based on extension theory. The self-adapting assessment function of the number of the clustering, clustering analysis using the FCM, and discrete process of the real features based on the maximum degree of membership were put forward. Subsequently, the degree of the importance for attributes from rough sets theory was introduced and the objective weights of features for the stored-rain pests were determined automatically. Finally, the familiar nine categories of the store-grain pests in grain-depot were recognized by a classifier based on the extension decision theory. The results show that the correct identification ratio is 93% and the rough sets weights application in the extension classification of the stored-grain pests is feasible.
參考文獻(xiàn)
相似文獻(xiàn)
引證文獻(xiàn)
引用本文
張紅濤,毛罕平.基于FCM離散化的粗集權(quán)重在糧蟲可拓分類中的應(yīng)用[J].農(nóng)業(yè)機(jī)械學(xué)報(bào),2008,39(7):124-128.[J]. Transactions of the Chinese Society for Agricultural Machinery,2008,39(7):124-128.