Competing Predictive Models for Covid-19 Data: a Case Study of Saudi Arabia

Almurashi, A. A.; Almarashi, A. M.; Khan, K.

The present study focuses on assessment of COVID-19 data viz-a-viz one with the study variable namely - daily confirmed cases in Saudi Arabia. Data for the present study covers the period from 15th March 2020 to 22nd October 2020. Prediction for the study variable was extracted using three time-series models. Accuracy of the competing models was assessed through mean absolute percentage error (MAPE). For visual representation of the data, relevant graphs are produced using forecast package of R software. Outcomes of the present study will have two pronged advantages. Firstly, the outcomes will provide administration of Saudi medical institutions an early warning system for devising effective coping strategies well in advance, for handling the ongoing pandemic and future recurrences (if any) of the same type. Secondly, relevant government officials dealing with such pandemics can formulate sustainable policies prior to any such recurrence in future.