Research Of Rule Extraction And Classification Algorithm Based On Neural Network

The whole world is facing the challenge of database blast and people often taste the puzzle of being submerged by data, but lacking knowledge with swift development of the Internet. The technology of data mining is a kind of solution. Data mining is a very important step of knowledge discovery in database, it is a course of distilling connotative, unaware and potentially useful information and knowledge from vast, incomplete, yawp, blurry and stochastic application data. The technology of data mining based on neural network solves the problems of data mining using the technology and advantages of neural network, so the technology of data mining can get better enhancement and perfect. Direct against the above situation, this dissertation has done the work of three following respects mainly.Firstly, the dissertation studies and analyses the basic concept, correlative techniques and theories of data mining and neural network. The part of data mining importantly studies the basic course of data mining, classification technology and late-model multidimensional data mining technology; The part of neural network importantly studies network study rules and frequently used neural network models.Secondly, this dissertation studies data mining course based on neural network in detail, and puts forward the cutting out and training algorithm and the optimization rule extraction algorithm of neural network, it also does the corresponding experiment and experiment result analysis aiming at the theory algorithm; This dissertation also studies the classification technology based on neural network, and puts forward the amelioration classification algorithm, it also does the corresponding experiment and experiment result analysis aiming at the classification algorithm; This dissertation ulteriorly studies the model of late-model multidimensional data mining based on neural network, and brings...