Research And Application On The Prediction Model Of Epidemic Situation In HFMD Disease Area Based On Big Data Analysis

HFMD(hand,foot and mouth disease)is an infectious disease caused by enterovirus infection.Since it was included in the national legal report of infectious diseases in 2008,it has become the Class C infectious disease with the highest incidence,and which prevention and treatment is of great significance.However,the traditional prediction model has poor prediction accuracy and low suitability,and is becoming increasingly unable to meet the needs of forecasting and warning.In recent years,with the rise of big data,artificial intelligence algorithms based on machine learning and deep learning have developed rapidly.Such algorithms have the ability of parallel processing and adjusting themselves,and these features make these algorithms suitable for fitting nonlinear problems that are affected by multiple factors and are unstable,which are exactly the problems in predicting the incidence of hand,foot and mouth diseases.Therefore,in theory,the application of machine learning technology in the accurate prediction of hand,foot and mouth disease will get better prediction effect.This paper is to explore the feasibility and application effect of this application.This acticle focuses on the application of machine learning technology in the prediction of hand,foot and mouth disease.In order to improve the prediction accuracy of hand,foot and mouth disease,the factors affecting the hand,foot and mouth disease epidemic are taken into account in the study.After determining the factors affecting hand,foot and mouth disease and quantifying the data,the data of hand,foot and mouth disease and the factors affecting the epidemic were obtained.After the data was processed,the model was established by using ARIMA(Integrated Auto Regression Moving Average Model),random forest and XGBOOST.For the problem that the XGBOOST algorithm consumes memory and is iteratively slow,the recursive feature elimination method is used to reduce the data dimension and improve system performance.After the highprecision prediction of hand,foot and mouth disease epidemic,the system function and non-functional requirement design,function design and database design were completed,and the big data analysis platform of hand,foot and mouth disease of Sichuan Provincial Disease Prevention and Control Center was realized and tested to improve hand,foot and mouth disease.The prediction accuracy of the epidemic situation provides a scientific basis for the prevention and early warning of hand,foot and mouth disease.MAE(Mean Absolute Error)and RMSE(Root Mean Square Error)were used as evaluation indicators.After comparison and analysis of three methods,the results showed that ARIMA was simple to construct and performed well in some areas;random forests performed steadily and were suitable for high processing.Dimensions,feature loss,unbalanced data,when the data is irregular,you can consider using this model;while the improved XGBOOST algorithm has higher memory consumption and higher resource requirements,but it is better than the other two algorithms in accuracy.The regional fitting effect is good,and the system is finally determined to use the model.It is believed that through the use of the big data analysis platform of hand,foot and mouth disease,the prediction of the opponent's foot and mouth disease will have a significant improvement in accuracy.