Machine Learning-Based Air Pollution Prediction

Abstract

Due to urbanization, air pollution (AP) has become a vital issue. AP is adversely impacting the humanity by causing asthma and other air-borne diseases. The accurate prediction of AP can aid in ensuring public health. Assessing and maintaining the quality of air has become one of the vital functions for the governments in several urban and industrial areas around the world today. There are a lot of parameters that play a significant role in increasing the air pollution like gases released from the vehicles, burning of remains of fuels, industrial gases, etc. So, with this increase in pollution, there is an increasing requirement of devising models which would record the information about the concentration of the pollutants. The deposition of the injurious air gases is adversely impacting people’s life, especially in urban areas. Machine learning (ML) is a domain that has recently become popular in AP prediction due to its high accuracy. There are, however, many different ML approaches, and identifying the best one for the problem at hand is often challenging. In this work, different ML techniques such as linear regression (LR), decision tree regressor (DTR), and random forest regressor (RFR) algorithms are utilized to forecast the AP. Results revealed that the RFR performed better than LR and DTR on the given data set.