Whether Prediction using Auto-correlation and Machine Learning Algorithms

As economic expansion accelerates, the deterioration of air quality persists. The primary catalysts for the initiation and progression of intense air pollution processes are the emissions of high pollution intensity and adverse weather conditions. The current scenario of rapid climate change has rendered traditional weather prediction methods ineffective and more challenging. Consequently, there is a pressing need for enhanced and dependable weather forecasting techniques. These forecasts have significant implications for a country’s economy and the well-being of its populace. The central goal of present study is assess and predict the air quality index (AQI) for Mumbai city of Maharashtra in India based on seasonal auto-regressive moving average (SARMA) with machine learning models. To validate the proposed technique, a six month dataset were utilized that were collected at the Shreemati Nathibai Damodar Thackersey Women’s University, Santacruz (w), Mumbai. This study provides the minimum root mean square error (RMSE) between predicted and recorded AQI. The performance comparative analysis of different kernels of support vector machine (SVM) with K-Nearest Neighbour (KNN) has been evaluated to validate the proposed method.