머신러닝 모델을 이용한 대한민국 코로나 신규 확진자 예측

Abstract

COVID-19, a coronavirus (COVID-19) caused 57,680 confirmed cases and 819 deaths in Korea as of December 28, 2020, causing many casualties worldwide. Predicting COVID-19 confirmed cases allows us to manage and plan effective preventive measures to reduce casualties. In this paper, we propose a methodology to predict COVID-19 new confirmed cases over the next four days using machine learning models. We propose using long short-term memory (LSTM), random forest, gradient boosting models. Experiments show that LSTM produces better prediction performance over other models in the majority of scenarios. We believe that this study is the first attempt to predict the trend of COVID-19 confirmed cases in Korea. We hope our work can inspire researchers to develop better methods to predict COVID-19 confirmed cases.