Machine learning model estimating number of COVID-19 infection cases over coming 24 days in every province of South Korea (XGBoost and MultiOutputRegressor)

Abstract

We built a machine learning model (ML model) which input the number of daily infection cases and the other information related to COVID-19 over the past 24 days in each of 17 provinces in South Korea, and output the total increase in the number of infection cases in each of 17 provinces over the coming 24 days. We employ a combination of XGBoost andMultiOutputRegressor as machine learning model (ML model). For each province, we conduct a binary classification whether our ML model can classify provinces where total infection cases over the coming 24 days is more than 100. The result is Sensitivity = 3/3 = 100%, Specificity = 11/14 = 78.6%, False Positive Rate = 3/11 = 21.4%, Accuracy = 14/17 = 82.4%. Sensitivity = 100% means that we did not overlook the three provinces where the number of COVID-19 infection cases increased by more than100. In addition, as for the provinces where the actual number of new COVID-19 infection cases is less than 100, the ratio (Specificity) that our ML model can correctly estimate was 78.6%, which is relatively high. From the above all, it is demonstrated that there is a sufficient possibility that our ML model can support the following four points. (1) Promotion of behavior modification of residents in dangerous areas, (2) Assistance for decision to resume economic activities in each province, (3) Assistance in determining infectious disease control measures in each province, (4) Search for factors that are highly correlated with the future increase in the number of COVID-19 infection cases. IntroductionThe new-type corona virus (SARS-CoV-2), which suddenly appeared at the end of 2019, caused global epidemic of COVID-19 in 2020. As of May 2020, industry, government, and academia from each country are cooperating to take measures to prevent the spreading of COVID-19 infection and develop therapeutic drugs. This is the third time in this century that humans have been threatened by corona virus. The first was SARS in 2003, the second was MERS in 2012, and the third is COVID-19. Comparing SARS and MERS with COVID-19, it can be said that computer science including artificial intelligence (AI) and machine learning (ML) has made great progress. In fact, many computer science approaches have been developed to prevent the spreading of COVID-19 infection. For example, AI and ML are used for infection spreading analysis, drug discovery assistance, automatic diagnosis, diagnosis assistance, social trend analysis, and infection route analysis. This paper proposes an ML technology for infection spreading analysis. To be more specific, we input the number of COVID-19 infection cases per day and other related information in each of 17 provinces in South Korea for the last 24 days, and output the total increase in the number of infection cases in each of 17 provinces over the coming 24 days. The result of conducting binary classification whether the total number of COVID-19 infection cases exceeds 100 over coming 24 days is Sensitivity = 3/3 = 100%, Specificity = 11/14 = 78.6%, False Positive Rate = ...

Applications of artificial intelligence in battling against covid-19: A literature review

“…Open source algorithms: In [474] , XGBoost is used to predict the number of infections in South Korea.…”

Hybrid Genetic Algorithm and Machine Learning Method for COVID-19 Cases Prediction

“…The same approach was conducted on the case of outbreak in Hungary [6], with a goal to show the potential of utilizing the machine learning technique in this domain. Machine learning was also used in [7] in order to estimate the number of reported cases in individual provinces of South Korea, by utilizing a combination of XGBoost and MultiOutputRegressor as a machine learning model.…”

Characterization of partially observed epidemics through Bayesian inference: application to COVID-19

“…Ref. [45] contains a comprehensive summary of various machinelearning methods used to "curve-fit" COVID-19 data and produce forecasts. Approaches that attempt to embed disease dynamical models into their forecasting process have also be explored, usually via compartmental SEIR models or their extensions.…”