Research On Prediction Model For COVID-19 And Data Visualization

In December 2019,a large number of people were infected in Wuhan City,Hubei Province,China,due to an epidemic outbreak caused by the COVID-19 virus.With the mutation of the virus and the passage of time,many countries have been affected by the COVID-19.With the outbreak of COVID-19 in countries all around the world,many countries are in a difficult time to fight the epidemic.Medical supplies and facilities are in short supply,and the development trend of the epidemic is difficult to be predicted intuitively.The World Federation of Engineering Organizations(WFEO)issued an appeal,calling on everyone to use AI and big data to propose corresponding solutions.Therefore,scientifically and effectively controlling the development of the epidemic is essential to do related prevention work and control.In face of the epidemic,many universities and research centers realize data visualization based on existing COVID-19 data,not only using existing infectious disease models but also construct some new models to predict the development trend of the epidemic.Existing models such as SI and Logistic models(non-linear quadratic models)are mainly used in epidemiology.The most common situation is to explore the potential risk factors of a certain disease,and then predict what is the probability of occurrence of the disease based on this factor.The SIR,SIRS,and SEIR models are systems of ordinary differential equations.The SEIR model is an infection model with an incubation period.Therefore,according to the characteristics of cases caused by COVID-19,this model can be considered to predict the development of the epidemic.In addition,in addition to realizing the SEIR model to predict the development of the epidemic,this paper will also compare,evaluate and screen several models,and select a suitable model from which to reasonably and effectively predict the future development trend of the epidemic.Among them,MLPRegressor is mainly used for the main prediction work.Therefore,the use of predictive models and data visualization part will be the research focus of this paper.In addition,except using multiple models for prediction,this article will also use the Echarts map plug-in to visualize the data,so that the data can intuitively reflect the development of the COVID-19 trend for not only china but other countries,and serve as a reference for epidemic prevention and control decisions and public behavior.