Simulation on Network Fault Diagnosis Based on RS-BPNN
Guo Jiang-ping · Jisuanji fangzhen · 2011
Network fault diagnosis problem is studied.Because network fault attributes have too many redundant,the traditional network fault diagnosis cannot delete the redundant attributes.To improve there fault diagnosis accurate,a fault diagnosis method is put forward based the BP neural network and rough set.Firstly,attribute sets are reduced by rough sets theory,then reduction attributes are used as the input variables of BP neural network,which accelerates the BP neural network training speed and improves the network fault diagnosis accuracy.The results show that the proposed method has fast convergence speed and high accuracy.