Font Size: 

The COVID-19 pandemic has spread across the globe. There are several countries that have finished experiencing a pandemic period, but there are even countries experiencing multi-wave epidemics. In phenomenological modeling based on the cumulative number of cases, a special method is needed if an area experiences a multi-wave epidemic. In this article, we use two changepoint detection methods, namely the pruned exact linear time (PELT) method and the interpolation method. Both methods produce changepoints in each of the existing waves where the PELT method gives more accurate results than the interpolation method. The changepoint results obtained by the PELT method are then used to determine the starting point for data collection for the last wave which is then used to make predictions using the Richards curve model. The prediction results obtained are in the form of the peak of the epidemic in an area and when it will end. So that it can provide input for the formulation of the government's medium-term strategy to overcome the pandemic that occurs and avoid a wave of repeated infections that are detrimental both in reducing the death rate due to COVID-19 and the economic condition of an area.