Abstract

During the first year of the Covid-19 pandemic, countries had no choice except Non-Pharmaceutical Interventions (NPIs) to control the spread of this infectious disease. Some of these interventions, like the stay-at-home order, were effective in controlling disease spread (Navazi et al., 2022); however, they caused adverse effects on the economic situation. So, we need a decision support system that can find effective NPIs in controlling the pandemic with low negative economic effects. This study aims to develop a multi-objective decision support system that can consider both. Since each pandemic has unique features, finding historical data to help decision-making is difficult, so the developed model should be able to consider existing evidence and best practices from other countries for providing suggestions (Navazi et al., 2024a). Since the effectiveness of NPIs is affected by society's adherence to NPIs, an indicator was added to the model to capture adherence with NPIs (Navazi et al., 2024b). The model should also consider pharmaceutical interventions after vaccine/cure development, because NPIs are still effective during the period it takes to reach herd immunity (Navazi et al., 2022). So, in this research, we used machine learning algorithms for learning from time series data gathered by Oxford University about NPI implementation levels. Moreover, a metaheuristic algorithm, an AI tool for optimization, is used to suggest the best level of NPI for the studied country based on evidence from other countries. The designed information system artifact is able to analyze time series data to discover knowledge about best practices and lessons learned for possible future pandemics.

Comments

tpp1415

Download

Share

COinS