Research and Application of Neural Network Based on Rough Set Theory

In order to analyze the performance of the neural network its black box characteristics, it is necessary to combine the neural network and the rough set theory, and use the rough set theory to initialize the parameters of the neural network, then determine, select and guide, finally put forward the application theory based on the neural network architecture of rough set. This theory is given various physical parameters. If the system outputs, after training the rough neural network, it can meet its requirements and performance, and there is a minimized error, which indicates that the experimental result is that the data mining task of the rough neural network can be completed by a better method. The accuracy values are verified and high in classification.