Long Term Forecasting of Ambient Air Quality Using Deep Learning Approach

With the rapid development of urbanization, ambient air pollution has become one of the most serious issues in the worldwide. Particulate matter particles have been found as one of the critical risk factors of several lung diseases and respiratory problems. Numerous countries worldwide are paying their close attention to these risk factors as it became more challenging day by day. Therefore, air quality assessment and forecasting is a major step to mitigate the environmental hazard. This paper proposes a time series based CNN-LSTM-SVR forecasting model, which can deal with the temporal dependency of the massive air pollution dataset and forecast air quality level over the next two weeks. It works better, approximately 91-96% than the baseline models. This model proved a better long term forecasting model, which is very reliable in the field of air quality modeling.