Bp Neural Network And Its Application In The Classification Of The Disease Prognosis

Artificial neural network have received much attention over the last few years, Back-propagation neural network belongs to mutilayered feed-forward networks, it adopts typical supervised learning, it is being used in the areas of prediction and classification and most useful in situations in which the relation between input and output is nonlinear. Properly trained network is able to extract any functional or statistical relationship between the input and output present in the training data. In this model , independent variable can be either continuous or discrete, it is not limited by strict assumptions of normality linearity variable independence etc. BP neural network is playing a important role as a non-linear model.In this paper, BP neural network is used for the classification of patients with primary biliary cirrhosis about survival, we adopt Levenberg -Marquardt algorithm, it makes learning time short, convergence fast; we use early-stopping method to avoid over-fitting and choose optimization model. In this study, we compare the performance of a neural network model and a Logistic regression model, we find that BP neural network gets good results in internal validation and external validation, so it is worthy tobe popularized, especially in the fields of prediction analysis and discriminate analysis.