Modified LSTM for Prediction of Pollutants Concentration

“Air pollution” is the term employed for those situations where airborne emissions are dangerous to the health of people or other living beings or when they negatively affect the environment or materials. A few typical causes of air pollution include cars, factories, homes with combustion gadgets, and forest fires. Carbon monoxide, particulates, nitrogen oxides, and sulphur dioxide are the contaminants that are most harmful to human health. Environmental and indoor air pollution, which causes respiratory illnesses and other disorders, has a major negative impact on morbidity and mortality. In this study, a unique prediction technique called Modified Long Short Term Memory (MLSTM) was created to prognosticate the daily measurements of pollutants such as SO₂, PM₁₀, PM_2.5, and NO₂ in Salem, Chennai, Thoothukudi, Madurai, and Coimbatore. In terms of predictability and execution time, the Modified LSTM approach’s performance was compared to the LSTM method. The Modified LSTM method performs better with higher prediction accuracy and lesser execution time.