Using self-organising maps to predict and contain natural disasters and pandemics

Area of Science:

• Epidemiology
• Public Health
• Data Science

Background:

• The COVID-19 pandemic underscored the need for effective predictive and containment strategies.
• Current responses focus on disease spread mitigation and vaccination, with future preparedness being crucial.
• Existing models may lack flexibility for diverse scenarios like pandemics and natural disasters.

Purpose of the Study:

• To propose a novel prediction and containment model for future pandemics and natural disasters.
• To develop a flexible, user-friendly data analytics tool for governmental and organizational decision-making.
• To demonstrate the model's efficacy in reducing infection rates while preserving economic activity.

Main Methods:

• Development of a prediction and containment model integrating selective lockdowns and protective cordons.
• Implementation of a data analytics model based on Self-Organising Maps (SOMs) for decision support.
• Comparative testing using publicly available data for Great Britain (GB).

Main Results:

• The proposed strategy effectively reduces peak infection rates.
• It enables a significant portion of regions (up to 25% of GB parliamentary constituencies) to maintain economic activity.
• The model facilitates rapid containment while allowing unaffected areas to operate normally.

Conclusions:

• The proposed model offers a viable strategy for managing future pandemics and natural disasters.
• Selective containment measures combined with data-driven decision-making can balance public health and economic stability.
• The Self-Organising Maps-based approach provides an accessible tool for effective crisis management.