Optimal Visual Predictive Modelling on Covid-19 Zones

Abstract

The COVID-19 pandemic has sparked an unprecedented global crisis, compelling nations to devise effective disease containment and mitigation strategies. Zoning, a critical aspect of public health management, categorizes regions based on COVID-19 severity levels, allowing targeted interventions. In this study, we propose an innovative predictive modelling approach leveraging advanced machine learning algorithms to anticipate and classify regions in India into red, orange, and green zones, reflecting varying levels of COVID-19 threat. Drawing from a vast dataset encompassing diverse socio-economic, demographic, and epidemiological variables. We develop predictive models using K-means clustering and decision tree classification techniques. The paramount significance of predicting COVID-19 zones is a key challenge here. By scrutinizing multivariate data, we identify significant factors influencing the spread and severity of the virus across different regions. This informed approach helps optimize resource allocation and public health interventions, allowing for a proactive response to potential outbreaks. The findings demonstrate the potential of machine learning in augmenting traditional epidemiological methods, aiding policymakers in making informed decisions. By continuously updating our predictive models with the latest data, we enable a dynamic and flexible zoning strategy, aligning with the current state of the pandemic. The projected methodology provides a foundation for developing robust decision-support systems, assisting healthcare authorities and policymakers in effectively navigating the challenges posed by COVID-19. The projected approach advocates for a holistic approach, encompassing healthcare infrastructure improvement and behavioural interventions to manage the pandemic effectively. The visual showing of affected areas with customized colours would report the information when the user interacts with the map.