AI-Powered Predictive Systems for Managing Epidemic Spread in High-Density Populations

Abstract

Predicting epidemic spread in high-density populations is a complex problem. This paper introduces a novel AI-based approach to predictive pathogen spread modeling in large airplane populations. It takes into account the complex interplay of short-term passenger mobility dynamics and AI-driven duration- and direction-based risk level estimations. A case study of airline passenger moving patterns is used to implement and validate prediction models. The originality of the proposed model consists of a specially adapted aggregating hierarchical clustering algorithm for building prediction models faster, thus increasing its practicability and scalability. The model is fully deployable and complies with major privacy standards. Preliminary results show good performance in evaluation. Due to its adaptive Bayesian nature, the model is easily extendable for other types of complex mobile settings, increasing the potential for wider application in critical infrastructure transportation security.

AI-Powered Predictive Systems for Managing Epidemic Spread in High-Density Populations. AI is increasingly used for predictive analytics in medicine, social sciences, biology, and related fields. AI-based approaches are particularly useful for implementing predictive systems for security purposes in critical infrastructure. The ability to predict, with high precision, potential risks and threats to large populations in insecure settings, and to do it on time, helps to develop measures for neutralizing these threats and ensuring safe functioning. The need for critical infrastructure security strategies is further enhanced by recent epidemics, which were not well managed, especially in transportation hubs. Various epidemic methods were proposed, but there is still a place and a need for innovative practical applications that are driven by real-world high-stakes issues.