Epidemic Analysis and Prediction of Novel Coronavirus based on XGBoost Algorithm

Abstract

The sudden outbreak of COVID-19 poses a great threat to the health and safety of people all over the world. Since the outbreak of the epidemic, the number of suspected and confirmed infections in the world has continued to rise. During the period of epidemic prevention and control, it is very important to carry out scientific research quickly, such as finding the source of the virus, controlling the spread of the virus, studying the pathogenic mechanism, collecting data, looking for scientific treatment and prevention and control programs, screening and developing effective drugs, and so on. At the same time, according to the data collected and published by the World Health Organization, it can reflect the response measures in different regions and the number of people found to be affected by COVID. The epidemic situation in COVID-19 is a critical period for major scientific research projects to tackle key problems, and an important stage to win time and save lives. Today, Covid is still rampant, and there are still different varieties. Based on the existing data, the COVID infection model can be calculated to predict the infectivity of new varieties, and different measures can be taken for countries with high mortality or infection rates. The experimental results show that, compared with three traditional machine learning algorithms, the model built based on the integrated algorithm XGBoost has the highest prediction accuracy rate for whether human beings are infected with novel coronavirus, and the accuracy rate reaches 95.6%, providing a non-destructive auxiliary method for biomedical detection of COVID-19.