Published Paper Details:

Accurate forecasting of air pollution, especially fine particulate matter (PM?.?), is crucial for protecting public health and guiding environmental policies. Traditional statistical models often struggle to capture the complex nonlinear and temporal patterns inherent in air quality data. This study introduces a hybrid model that integrates Long Short-Term Memory (LSTM) deep learning networks with the Particle Swarm Optimization (PSO) algorithm to enhance the prediction accuracy of PM?.? concentrations. LSTM networks are well-suited for modeling sequential time-series data due to their ability to retain long-term dependencies, while PSO efficiently optimizes hyperparameters to improve model performance. The proposed LSTM-PSO model was evaluated using extensive real-world air quality datasets collected from major urban centers over multiple years. Results demonstrate that the hybrid model significantly outperforms standalone LSTM and traditional machine learning approaches, achieving lower Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). Moreover, the integration of PSO not only improved prediction accuracy but also accelerated the convergence speed of the LSTM training process. These findings highlight the effectiveness of combining deep learning with metaheuristic optimization algorithms for robust and efficient air quality forecasting, offering valuable insights for environmental monitoring and public health management.