[논문 리뷰] SEIR and Regression Model based COVID-19 outbreak predictions in India

Gaurav Pandey, Poonam Chaudhary외 2명|arXiv (Cornell University)|2020. 04. 01.

COVID-19 epidemiological studies참고 문헌 11인용 수 118

한 줄 요약

The paper compares SEIR and regression models to predict short-term COVID-19 cases in India (Jan 30–Mar 30, 2020) and forecasts 5,000–6,000 cases in the next two weeks, evaluating with RMSLE and estimating R0.

ABSTRACT

COVID-19 pandemic has become a major threat to the country. Till date, well tested medication or antidote is not available to cure this disease. According to WHO reports, COVID-19 is a severe acute respiratory syndrome which is transmitted through respiratory droplets and contact routes. Analysis of this disease requires major attention by the Government to take necessary steps in reducing the effect of this global pandemic. In this study, outbreak of this disease has been analysed for India till 30th March 2020 and predictions have been made for the number of cases for the next 2 weeks. SEIR model and Regression model have been used for predictions based on the data collected from John Hopkins University repository in the time period of 30th January 2020 to 30th March 2020. The performance of the models was evaluated using RMSLE and achieved 1.52 for SEIR model and 1.75 for the regression model. The RMSLE error rate between SEIR model and Regression model was found to be 2.01. Also, the value of R0 which is the spread of the disease was calculated to be 2.02. Expected cases may rise between 5000-6000 in the next two weeks of time. This study will help the Government and doctors in preparing their plans for the next two weeks. Based on the predictions for short-term interval, these models can be tuned for forecasting in long-term intervals.

연구 동기 및 목표

• Analyze the COVID-19 outbreak in India up to March 30, 2020.
• Develop short-term forecasts for the next two weeks using SEIR and regression models.
• Evaluate and compare the predictive performance of the two models.
• Estimate the basic reproduction number R0 for India during the study period.

제안 방법

• Use SEIR and regression models for predictions calibrated on data from Jan 30, 2020 to Mar 30, 2020.
• Data sourced from the Johns Hopkins University repository.
• Evaluate model performance using RMSLE (SEIR = 1.52; regression = 1.75).
• Compare models with RMSLE difference of 2.01 between SEIR and regression predictions.
• Estimate R0 for the spread of the disease (R0 = 2.02).
• Provide short-term forecasts suggesting 5,000–6,000 expected cases in the next two weeks.

실험 결과

연구 질문

• RQ1 Can SEIR and regression models accurately forecast short-term COVID-19 case counts in India during the study window?
• RQ2 How do SEIR and regression models compare in predictive performance (RMSLE) for India’s COVID-19 data?
• RQ3 What is the estimated basic reproduction number (R0) for India in the study period?
• RQ4 What are the projected case ranges for the upcoming two weeks based on these models?

주요 결과

• SEIR model achieved RMSLE of 1.52.
• Regression model achieved RMSLE of 1.75.
• RMSLE difference between SEIR and regression models is 2.01.
• Estimated R0 for the outbreak is 2.02.
• Short-term forecasts predict 5,000–6,000 additional cases in the next two weeks.