Air Quality Prediction: Big Data and Machine Learning Approaches

Abstract

Abstract-Monitoring and preserving air quality has become one of the most essential activities in many industrial and urban areas today. The quality of air is adversely affected due to various forms of pollution caused by transportation, electricity, fuel uses etc. The deposition of harmful gases is creating a serious threat for the quality of life in smart cities. With increasing air pollution, we need to implement efficient air quality monitoring models which collect information about the concentration of air pollutants and provide assessment of air pollution in each area. Hence, air quality evaluation and prediction has become an important research area. The quality of air is affected by multi-dimensional factors including location, time, and uncertain variables. Recently, many researchers began to use the big data analytics approach due to advancements in big data applications and availability of environmental sensing networks and sensor data. The aim of this research paper is to investigate various big-data and machine learning based techniques for air quality forecasting. This paper reviews the published research results relating to air quality evaluation using methods of artificial intelligence, decision trees, deep learning etc. Furthermore, it throws light on some of the challenges and future research needs.Index Terms-Air quality evaluation, big data analytics, data-driven air quality evaluation, and air quality prediction.

Environmental Health Surveillance System for a Population Using Advanced Exposure Assessment

“…By using the machine learning approach, Zheng et al [32] developed a U-Air system that combines different types of heterogeneous big data to estimate air quality, such as meteorology, traffic flow, human mobility, road network structure, and point of interest. Recently, many researchers have begun to use the big data analysis approach because of the development of big data applications and the availability of environmental detection networks and sensor data [68]. Air quality has been estimated by a deep learning and image-based model [69].…”

Prediction and Optimization of Air Pollution-A Review Paper

“…Decision Tree Model comes under the Supervised Learning Technique [10]. It is collection of branches and nodes, where each node represents the various choices based on the observations and the output or final decision is represented by the leaf nodes.…”

“…They proposed a model which can predict hourly prediction of pollutants using a multi task learning (MTL) problem and also used various normalization techniques which helped to achieve better performance and accuracy rate. Gaganjot Kaur Kang, Jerry Zeyu Gao, Sen Chiao, Shengqiang Lu, and Gang Xie presented a review paper [10] on the big data analytics approaches and machine learning for forecasting the air quality index. According to the paper [4] it has made a research work on predicting PM 2.5 concentrations using RNN (Recurrent Neural Network) with LSTM (Long Short Term Memory) because it has been found that PM 2.5 has been severely effecting the human health and much research has been already proposed for predicting it so here RNN along with LSTM is used, which is a high level Neural Network API written in python.…”

Applicability of machine learning in modeling of atmospheric particle pollution in Bangladesh

“…Though the models do not unambiguously simulate the environmental process, in general, they exhibit better prognostic performance than the CTMs on spatiotemporal scale in the existence of extensive monitoring records (Marshall et al 2008 ). Several studies have been conducted in different countries to evaluate the performance of machine learning models in the field of air quality modeling and forecasting (Kang et al 2018 ). However, based on relevant literature, the study of machine learning in air pollution modeling was limited in Bangladesh, though multiple studies were performed to investigate the particulate pollution (Begum et al 2011 ; Begum and Hopke 2018 ).…”