Analysis Of Time Series Based On Neural Network In Type ? Diabetes

With the development of social economy and the improvement of living standards,type ? diabetes has become an important disease that seriously threatens human health of body and mind.A large number of studies have shown that community intervention in type ? diabetes is recognized as a cost-effective and effective measure to reduce cardiovascular morbidity in the world.Therefore,the study evaluates the risk of individuals with diabetes and then takes effective intervention measures could delay the occurrence of disease and reduce the harm.The traditional time series prediction method can not analyze and fit the trend of high non-linear and multifactorial influence,which has the problems of low prediction accuracy and slow prediction.Artificial neural network(ANN)is one of the intelligent pattern recognition tools currently,with its good nonlinear approximation and the adaptive,self-organizing characteristics.These charaterisics have been gradually applied in the field of medical auxiliary.The application of the classification and diagnosis of infectious diseases,cancer,high blood pressure and related diseases has been reported.And the BP neural network had become the most popular time series modeling and forecasting method.The characteristicof the BP neural network model is that it can approximate any continuous function with arbitrary precision infinite,has strong nonlinear mapping ability,generalization ability and fault tolerance.With type ? diabetes several important indexes for the research data as the foundation,explore the role of the BP neural network prediction in diabetes,then combine with the traditional time series ARMA model aalysis the risk for the future situation,discuss the individual best forecasting model of type ? diabetes risk.Finally,through an example analysis of hospital data,the model is analyzed to verify that the model can predict the future risk of the individual.