Track: COVID-19 Analytics Competition

Abstract

Large-scaled distribution of vaccines can be highly complex and dynamic. This paper aims to propose an integrated framework that addresses the limitations of former literature in optimizing vaccine allocation for controlling the covid-19 outbreak in Qatar. First, we predicted the total positive cases in Doha for the upcoming 14 days using Autoregressive Integrated Moving Average (ARIMA) modeling. The best fit model was ARIMA (4,2,4) based on BIC (Bayesian Information Criteria) with an overall MAPE of 4.86% and R2 of 0.9973 values. Then, we formulated a mathematical model to optimally allocate covid-19 vaccines to the twenty-five Primary Healthcare Centers (PHCCs) considering the total associated disease spread risk among the population, operational capacity limitations, and the transfer of ATP (Available to Promise) quantities between the centers. The obtained results provided managerial insights on how decision-makers can create efficient logistical capabilities for covid-19 vaccine allocation.