Portable Air Quality Monitoring System in ANN Using Combination Hidden Layer Hyperparameters

Abstract

The transportation and industrial sectors are growing rapidly, with positive and negative consequences in the form of air pollution. According to the Global Alliance on Health and Pollution (GAHP), 3.4 million people died from air pollution-related causes worldwide in 2017, with 123,700 of them dying as a result of air pollution. As a result, a portable system was built in this study to monitor air quality and categorize it using the Artificial Neural Network (ANN), with the classification results displayed on an Android application. Air quality classification is accomplished by varying the hyperparameters of the Artificial Neural Network (ANN), such as the number of hidden layer neurons, dropout, and batch size, while utilizing the gas parameters PM10, PM2.5, NO2, SO2, CO, and 03. The classification results will also be classified into five categories: good, moderate, satisfactory, poor, and very poor air quality. The system is intended to give accurate results.