Air pollution has become one of the most critical environmental and public health challenges worldwide, particularly in rapidly urbanizing regions. Accurate and real-time prediction of air quality is essential for enabling timely preventive measures and improving overall quality of life. This project, titled "AirSense — Live Pollution Level Predictor", presents an intelligent and interactive system designed to predict Air Quality Index (AQI) using machine learning techniques.

The system utilizes a Random Forest Regression model trained on a realistically simulated multi-city dataset inspired by global pollution patterns. The dataset incorporates key environmental and atmospheric parameters such as PM2.5, PM10, NO₂, O₃, CO, SO₂, temperature, humidity, wind speed, precipitation, time, and seasonal variations. By capturing both spatial and temporal pollution characteristics, the model achieves reliable AQI predictions with strong performance metrics.

AirSense integrates real-time geolocation capabilities, allowing users to automatically detect their location and obtain localized air quality predictions. The system provides an interactive web-based interface built using Flask, where users can manually adjust pollutant levels through dynamic sliders and instantly visualize AQI predictions. Additional features include health impact analysis, feature importance visualization, city-wise comparison, and short-term AQI forecasting.

The model is evaluated using standard metrics such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and R² score to ensure prediction accuracy and robustness. Overall, this project demonstrates how machine learning and web technologies can be combined to create an effective decision-support system for environmental monitoring.