Home // SMART 2021, The Tenth International Conference on Smart Cities, Systems, Devices and Technologies // View article

A Data-Driven Approach for Region-wise Environmental Health and COVID-19 Risk Assessment Scores

Authors:
Sanjana Pai Nagarmat
Saiyed Kashif Shaukat

Keywords: COVID-19, air quality index, tree cover, health score, risk score

Abstract:
High population density in India has led to rapid influx of citizens into urban cities. With urbanization, cities are facing massive issues in terms of improving citizen health and living conditions, medical infrastructure facilities and overall environmental conditions. The recent COVID-19 pandemic further threw light on the need to solve these pre-existing challenges. Academic scholars, researchers and industries across the country have developed several platforms and smart city applications to help city planners. However, the data sources largely remain segregated. Due to this, applications have restricted ability to provide interpretable results and quantified scores at a granular level. Our paper introduces a novel data-driven approach that helps city planners easily comprehend the current situation and further prioritize action plans to solve urban challenges of environmental health and medical infrastructure facilities in city wards. Data from several sources are combined to provide scores at a granular ward level that is indicative of ward environmental health and ward level risk in situations like the pandemic. Health score provides recommendations to improve the environmental health while the risk score helps in identifying critical zones and predicting the number of active patients in the region. Several Machine Learning based scoring models to predict risk scores for wards are considered. Risk scores are analyzed on a spatio-temporal basis and results are validated with the actual data from Pune smart city. The model introduced in this paper based on Gradient Tree Boosting Algorithm successfully predicts the ward risk scores with 94.55% accuracy.

Pages: 14 to 23

Copyright: Copyright (c) IARIA, 2021

Publication date: May 30, 2021

Published in: conference

ISSN: 2308-3727

ISBN: 978-1-61208-805-1

Location: Valencia, Spain

Dates: from May 30, 2021 to June 3, 2021