A Comparative Assessment SARIMA and LSTM Models for the Gurugram Air Quality Index's Knowledge Discovery

Gurugram, an urban center in India, experiences elevated levels of air pollution as a result of rapid urbanization and intensified industrial operations. Air pollution, responsible for 25% of all deaths in underdeveloped countries, can lead to long-term health conditions such as lung cancer and cardiovascular disease. Precise forecasting of the air quality index (AQI) is crucial for the implementation of effective environmental and public health measures. The objective of this project is to evaluate the precision of Gurugram's Air Quality Index (AQI) predictions by employing two machine learning algorithms, namely SARIMA and LSTM. This study aims to gain knowledge discovery by comparing the accuracy of SARIMA and LSTM models in forecasting AQI for Gurugram that can be utilized for safeguarding public health and detecting potential emergencies at an early stage. Discoveries indicate that although the LSTM model has a higher degree of error variability (with a root mean square error of 63.163 compared to 61.999 for SARIMA), it surpasses SARIMA in terms of average forecast accuracy (with a mean absolute error of 40.948 versus 45.758).