Research and Prospect of Early Warning and Prediction Model of Public Health Emergencies Based on Big Data Computation

The systematic review method was used to analyze the distribution of literature for understanding the trend of early warning methods. Regression analysis using spatiotemporal models such as ARIMA models and propagation dynamics models was used in this study. Geographic visualization techniques and neural networks, and even traditional statistical models were combined with big data and geographic information systems for early warning. 72 articles were reviewed, and the result showed that public health emergencies have gradually changed from traditional models In practical applications, it is necessary to comprehensively consider the characteristics of the disease, data types, and model characteristics to select the most appropriate predictive model. The model can be compared with a variety of methods to provide technical support.

I. Introduction

Before the occurrence of an event and when the evidence of causality and reaction relationship has not been determined, the area, scale, nature, influencing factors, radiation range, harm degree, and possible consequences of the event can be comprehensively evaluated based on the collected information and intelligence data. Appropriate means must be adopted to issue warnings of event threats in advance, and corresponding actions must be taken to prevent the occurrence and development of events to the maximum extent [1]. Early warning of public health emergency refers to the warning signal issued before or in the early stage of a public health event to remind of the possibility of an outbreak or epidemic of a public health event or the risk that the scope of its occurrence may expand [2].