Epidemic Prediction Of Infectious Diseases Based On Time Series Model

In recent years of infectious diseases endemic to the development of the world caused serious harm, from China’s SARS to African Ebola, the infectious disease control and prediction has become a hot issue in the scientific research, use the scientific technology and methods to control the development of infectious diseases, reduce impact of infectious disease is crucial for social development. In this paper, the epidemic trend of infectious diseases and forecasting methods are studied, the main work is as follows: 1. missing data processing. This paper is mainly based on the time series model for the prediction of the epidemic trend of infectious diseases, so the sample data is a time series, which contains the statistical results of the number of infectious diseases in different time periods. But due to the incomplete statistics, the inevitable emergence of some period of time in the statistical missing data, this paper first on the principle of data preprocessing and commonly used method were introduced. Secondly, the cubic spline interpolation method were introduced, and based on the cubic spline interpolation of the sample data are processed. 2.the introduction of prediction model. The classical model to predict the extrapolation of data this paper classified summary of common model principle and method are introduced, which focuses on time series model. Prediction of epidemic trend of ARIMA time series model, based on the 3.This paper selects the class B infectious disease hepatitis, tuberculosis and influenza C infectious disease as the research object, collecting time series data, the lack of data processing, the calculated sample data required, based on the model between different and different types of infectious diseases prediction benchmarking results there, trend of class B infectious diseases of seasonal obvious and self correlation, so the model has relatively good prediction results, and for influenza, the epidemic trend and there is no obvious periodic law and self correlation, the prediction results are not ideal, so the study on the prediction of infectious diseases and the need for specific diseases, selecting the appropriate model. 4.prediction of epidemic tendency of infectious diseases based on ARIMA model. ARIMA product model for time series is periodic, seasonal variation is more accurate, and the model can describe the time series data of randomness, periodicity, seasonal variation. ARIMA model based on the time series of the number of hepatitis B prevalence were predicted as the object, the results show that the prediction effect is better than the ARIMA model. Finally, the summary and outlook of the work done in this article and on the prediction of the development of infectious diseases.