AiCareAir: Hybrid-Ensemble Internet-of-Things Sensing Unit Model for Air Pollutant Control

The detrimental effects on human health caused by air pollution show that being able to predict air quality is a task of utmost significance. The application of artificial intelligence (AI) and the Internet of Things (IoT) is seen as promising in this domain. The performances of state-of-the-art models in terms of prediction accuracy vary with different pollutants and are acceptable only for certain pollutants. This article uses machine learning (ML) and deep learning (DL) models to predict the concentrations of six major air pollutants. Data are collected over eight months with 1400 daily instances from sensors deployed in Kuala Lumpur, Malaysia. As an intelligibly robust system, in this article a hybrid-ensemble model is proposed using a combination of ML models, specifically random forest, K-nearest neighbor (KNN), extreme gradient boosting (XGBoost), and neural network (NN) models, namely, long short-term memory (LSTM), gated recurrent units (GRUs), and convolutional NNs (CNNs). Here, a hybrid-ensemble learning model is created using five various ML models as weak learners. In previous ensemble models, a homogeneous group of weak learners are used; however, this work uses a heterogeneous group of weak learners. The prediction accuracy is compared using $R2$ score, absolute, squared, and root-mean-squared errors (RMSEs).

I. Introduction

Current global environments face a major challenge in air pollution of which a substantial portion remains undetectable to the human eye. Worldwide, national agencies have established air quality standards for some of these pollutants to protect public health. Six contaminants have been classified as criteria air pollutants by the Environmental Protection Agency (EPA) because higher levels may have an impact on human health and/or there are environmental-based criteria (science-based guidelines) for assessing acceptable levels. These six air pollutants are carbon monoxide (CO), nitrogen dioxide (NO2), surface ozone (O3), sulfur dioxide (SO2), and particulate matters (PM2.5 and PM10). Accurate estimation of these major pollutants has significant environmental and health implications. These estimations can empower policymakers, environmental agencies, and communities to take proactive measures to reduce pollution levels, minimize health risks, and protect ecosystems. Moreover, accurate prediction can facilitate the development of sustainable practices, cleaner technologies and more effective regulatory frameworks to improve air quality and safeguard human and environmental well-being.