Imputation of new COVID-19 cases missing data using basic statistical methods

This study investigates the imputation of new COVID-19 cases with missing data in Kedah and Selangor states by using basic statistical methods. This study aims to impute missing data using four basic statistical methods and compare the methods using a performance index. The four basic statistical methods applied in this study are Linear Interpolation, Top Bottom Average, 7-Day Average, and 14-Day Average. The time series data employed is the number of new COVID-19 cases for 365 days, which was recorded in daily numbers for 2021 in Kedah and Selangor states. The time series data was sampled and randomly removed by 10% and 20%. The removed data will be imputed using the four basic statistical methods. The performance indices used to compare the performance of the basic statistical methods were mean absolute error (MAE), root mean square error (RMSE), and correlation coefficient (CC). Overall, Linear Interpolation and 14-day Average are suitable basic statistical methods for finding missing data. The findings of this study suggest that basic statistical techniques can be instrumental in supporting the Malaysian Ministry of Health (MOH) in filling gaps in data on new COVID-19 cases in future initiatives.