Air Quality Index forecasting using parallel Dense Neural Network and LSTM cell

摘要

Air pollution is a growing threat towards society and various measures are being taken recently to control it. The problem of concern which remains is the efficient prediction of air pollution to work in the right direction for reducing the same. Since the AQI follows a periodic pattern, deep learning models can be used to effectively predict the future AQI values. LSTM being a prominent time series forecasting model can be integrated with a separate DNN model to effectively add the impact of weather, temperature and other factors that can affect the future AQI values. The paper also explores the impact of having a parallel DNN to the LSTM cell instead of using the cell alone.