On Predicting COVID-19 Fatality Ratio Based on Regression Using Machine Learning Model

The world has been in the grips of the Coronavirus Disease-19 (COVID-19) pandemic for almost two years since December 2019. Since then the virus has infected over a hundred and fifty million and has resulted in over three million deaths. However, fatality rates have been observed to be drastically different in different countries. One reason could be the emergence of variants with differing virulence. Other factors such as demographic, health parameters, nutrition levels, and health care quality and access as well as environmental factors may contribute to the difference in fatality rates. To investigate the level of contributions of these different factors on mortality rates, we proposed a regression model using deep neural network to analyze health, nutrition, demographic, and environmental parameters during the COVID-19 lockdown period. We have used this model as it can address multivariate prediction problems with higher accuracy. The model has proved very useful in making associations and predictions with low Mean Absolute Error (MAE). © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Consequences of lockdown caused by COVID-19 outbreak on the quality of air in Dhaka

Air pollution has become a worldwide problem that has a negative impact on both human health and the environment. The development of systems for predicting air pollutant severities ahead of time is being driven by this rising threat. In this paper, we proposed using a Long short-term memory (LSTM) model of an Artificial Neural Network (ANN) to predict air pollutant severity levels, as a time series during the COVID-19 lockdown period, providing an early warning. The research used three types of real time datasets of Dhaka city that included records of three gaseous pollutants (CO, NO2, PM2.5). Modeling of the dataset of each pollutant was carried out on hourly and minute-based intervals in two different locations, Mirpur and Baridhara. The predicted results were compared with the readings of the dataset and the model attained high accuracy in predicting air quality. Finally, the air pollutants data were analyzed with COVID 19 cases. Our analysis reviews that the concentrations of air pollutants are in agreement with the regional COVID 19 cases. © 2021 IEEE.

Monitoring Air Quality of Dhaka using IoT: Effects of COVID-19

Air pollution and COVID-19 both are the most provocative issue nowadays. Air pollution holds a dangerous impact on the COVID-19 issue as well as human health. The urban cities like Dhaka are under stress to remain habitable. With the huge density of transportation and population, air quality index is to be monitored minutely and the impact of the COVID-19 pandemic is to be observed. It is required to develop an Internet of Things based remote monitoring system to observe the air particularity in the different areas of the Dhaka city and make a comparison between the before and during the COVID-19 Pandemic. The platform aims to track out the concentration of gases in the Dhaka City like carbon monoxide (CO), nitrogen di-oxides (NO2) on real-time that provides air quality index (AQI). Using Arduino based Node MCU and the sensors are to detect substantive conditions of gases. ESP-32 Wi-Fi module is used to send the data to the server so that it can be accessed from anywhere. The data is taken before and during COVID-19 period with the developed IoT platform. It shows the difference between the CO and NO2 emission on the impact of the pandemic with the measured value. This study will help us to make further decisions and action regarding air pollution. © 2021 IEEE.