AI-POWERED MULTI-CITY AQI FORECASTING SYSTEM
DOI:
https://doi.org/10.62643/ijerst.2026.v22.n1(2).pp308-312Keywords:
ir Quality Index; XGBoost; LSTM; Time-Series Forecasting; Feature Engineering; Lag Features; Multi-City; Multilingual; Accessibility; StreamlitAbstract
Air pollution poses a critical public health challenge in India, affecting millions of citizens across 29 major cities. This paper presents an AI-powered multi-city Air Quality Index (AQI) forecasting system that integrates an XGBoost regression model with lag-based time-series feature engineering on a dataset of 841,812 daily records (2022–2025). The primary XGBoost model achieves an R² score of 95.07%, MAE of 2.972, and RMSE of 24.900 on the held-out test set. A multilingual Streamlit interface supports English, Hindi (ह िंदी), and Telugu (తెలుగు), delivering colour-coded AQI categories, health advisories, and bilingual voice alerts via Google TTS and a pyttsx3 offline fallback, making accurate air quality insights accessible to diverse and non-literate user populations
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
