🌬️ AQI Prediction System

Overview

In an era of rising urbanization and industrialization, air pollution has become a critical global health concern. This project provides a comprehensive end-to-end solution for predicting the Air Quality Index (AQI) using machine learning.

By analyzing the concentration of key pollutants (PM2.5, PM10, NO2, CO, etc.), the system provides high-precision forecasts, helping citizens and authorities make informed decisions. It bridges the gap between raw environmental data and actionable health insights through a modern, high-performance web interface.

Features

• 🚀 Multi-Model Prediction: Choose between Random Forest, XGBoost, and Linear Regression for tailored accuracy.
• 📊 Granular Pollutant Analysis: Processes 11 distinct parameters including PM2.5, PM10, NOx, CO, O3, and Benzene.
• 🎨 Modern UI/UX: A stunning dark-mode dashboard with real-time feedback and glassmorphic design elements.
• ⚡ Fast Inference: Optimized model serialization using `pickle` for near-instantaneous results.
• 📍 City-Specific Insights: Context-aware forecasting based on localized historical data.

Pipeline & Architecture

How It Works

1. Data Ingestion: User inputs pollutant levels through the UI.
2. Preprocessing: Data is converted to a NumPy array; if Linear Regression is chosen, a pre-trained scaler is applied.
3. Model Inference: The selected `.pkl` model predicts the raw AQI value.
4. Classification: The raw value is categorized into status levels (Good, Moderate, Poor, etc.) based on standard environmental benchmarks.
5. Visualization: The UI dynamically updates the AQI scale and color-coded results.