Feasibility of Multilayer Perceptron (MLP) Network to Correlate Air Quality Index (AQI) and COVID-19 Daily Cases

• Published in: Springer Proceedings in Physics
• Main Author: Abd Maruzuki M.I.F.; Tengku Zahidi T.S.A.; Khairudin K.; Osman M.S.; Jasmy N.F.; Abdul Hadi B.; Akbar M.S.; Saufie A.Z.U.; Fathullah M.; Nor Syamsudin D.S.; Mohd Nazeri N.B.
• Format: Conference paper
• Language: English
• Published: Springer Science and Business Media Deutschland GmbH 2023
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85168766496&doi=10.1007%2f978-981-19-9267-4_43&partnerID=40&md5=b46979f5e1b975b3f5ae1cd182bc4470

A movement control order (MCO) was implemented in Malaysia starting from March 18th, 2020, as a pandemic control strategy that restricted all movement and daily outdoor activities to curb the transmission of COVID-19 pandemic. The most populated area in Malaysia, Petaling Jaya, Selangor, was selected to investigate the relationship between the COVID-19 outbreak and air pollution. Multilayer perceptron (MLP) model was used in this study to correlate air quality index (AQI) with COVID-19-related cases/deaths. The underlying hypothesis is that a pre-determined particulate concentration can encourage COVID-19 airborne transmission and make the respiratory system more susceptible to this infection. The in-silico strategy employed an innovative machine learning (ML) methodology, specifically MLP network using AQI data from the Department of Environment (DOE), Malaysia as input data and number of COVID-19 cases from the Ministry of Health, Malaysia as target data. The MLP model was trained for 10,000 times. Based on the results obtained, the model starts to converge after 1000 epochs with a small mean absolute error (MAE) (173.0–118.9) from day 1 to day 14. However, there is no definitive correlation between predicted COVID-19 patients and the AQI with respect to day configuration. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.