Intelligent Teaming in Public Health Management

In the context of decision-making in public health settings amidst uncertainties, the effective fusion of human intelligence and artificial intelligence (AI) plays an important role. In this research paper, we introduce the CI framework as a solution to effectively incorporate the two forms of intelligence via a layered structure comprised of predictive modeling, explainability, and decision validation by experts. Our proposal includes a confidence-weighted fusion technique that leverages the confidence values assigned by both humans and machines to facilitate effective collaboration between the two. Based on our experiments performed based on empirical data from epidemiology cases pertaining to COVID-19, our proposed CI technique has the advantage of offering improved accuracy, robustness, and consistency of decisions compared to the human-only and AI-only baseline approaches. Finally, our work also bridges the gap between theoretical formulation and implementation. Although we observe promising performance gains, there are some limitations in terms of expert simulation.