Forecasting model of Covid-19 cases using fuzzy time series using persentage change

Husain, A. A.; Surarso, B.; Farikhin, Irawanto B.

The pandemic caused by the novel corona virus (covid-19) has affected various aspects of life throughout the world. Indonesia is one of the country with a daily high number of Covid-19-19 spread cases. This study aims to obtain a forecasting model of Covid-19 cases that can be used to predict Covid-19 cases daily and it can increase the readiness of Covid-19 health protocols system. In this study, we get a very good model for Covid-19 forecasting in Indonesia obtained by the fuzzy time series method using frequency density-based partitioning. The universe of this method is the percentage of case changes from day to day. The percentage change as a universe in fuzzy time series forecasting method makes the results of comparison of actual data and predictions increasingly similar. We use data of the Covid 19 cases taken from the Nasional Kompas website during June 2020. Forecast results show very good with MSE value of 457,83 and small AFER value of 0,0425138%.