Real Time Noise Pollution Prediction in a City using Machine Learning and IoT

This study introduces a novel approach to tackle urban noise pollution in Vijayawada city by leveraging Machine Learning (ML) and Internet of Things (IoT) technologies. The aim is to predict real-time noise pollution levels, with a particular emphasis on the impact of road traffic. The methodology integrates four regression models (Decision Tree, Gradient Boosting, Support Vector, Linear Regression with Grid Search) and three forecasting models (ARIMA, Holt-Winters Exponential Smoothing, SARIMA), utilizing data collected from IoT acoustic sensors across diverse landscape configurations. This research underscores the importance of ML and IoT in addressing urban noise pollution challenges and implications for urban planning and policy making, acknowledging both strengths and limitations of the proposed approach. In Existing models there is no real time data used for prediction. But this case study includes real time data collected from IOT sound sensor which give better results.