PREDIKSI JUMLAH KASUS COVID-19 MENGGUNAKAN METODE LINEAR REGRESSION DAN POLYNOMIAL REGRESSION

Abstract

T Coronavirus Disease 2019 (Covid-19) is an infectious disease caused by anew type of coronavirus called SARS-CoV-2 (Severe AcuteRespiratory Syndrome Coronavirus-2). This disease primarily attacks the human respiratory systemand can cause symptoms ranging from mild, such as fever and cough, to severe, such as pneumonia andacute respiratory syndrome. Initial transmission occurred from animals to humans andis transmitted between humans through droplets, direct contact, or contaminated surfaces. In this study, the main focus is on usingLinear Regression and Polynomial Regression algorithms to predict thenumber of Covid-19 cases occurring in Indonesia. The data usedis historical data on the percentage of the spread of Covid-19 cases inseveral regions in Indonesia and is of a time series nature. The results obtainedfrom this study are that the Linear Regression method is thebest method compared to Polynomial Regression because the evaluation resultswith R2-square show that this method has a percentage of 95.56%greater than the other methods used, which only reached 93.76%. In addition, it can be concluded that this is because the value of the evaluation results using theRoot Mean Square Error (RMSE) method, the Polynomial Regression method obtained avalue of 1373.25 and the Linear Regression method only 1043.94,so that the best performance in predicting the number of Covid-19 cases is theLinear Regression algorithm.