Air Quality Prediction using Recurrent Air Quality Predictor with Ensemble Learning

Air Pollution is a key concern for countries all over the world. It is hazardous to the health of people and the environment; more importantly, it is mostly invisible. It is important to consider how much is their concentration at a given time so that the appropriate action may be taken. There are several methods to predict future air pollutant concentration, namely the recurrent air quality predictor, but its prediction accuracy falls off after a few hours. Hence, this paper aims to create a device that can predict future air pollutant concentrations through an ensemble of machine learning algorithms, RAQP included. After creating and testing the ensemble model, statistical tests on the mean PLCC and RMSE of the ensemble model for the first 12 time step predictions yielded significantly better results compared to the RAQP.