AI-Driven Air Pollution Forecasting Using Machine Learning: A Case study in Ulaanbaatar

Sosorburam, Uuganbolor (2025) AI-Driven Air Pollution Forecasting Using Machine Learning: A Case study in Ulaanbaatar. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (3MB) | Preview

Abstract

Air pollution in Ulaanbaatar, Mongolia, poses a major public health risk, with PM2.5 levels frequently exceeding WHO limits. This study proposes a machine learning–based framework to forecast short-term PM2.5 concentrations using publicly available air quality and weather data. Hourly pollution data from the OpenAQ API was combined with meteorological data from Weather.com to create a time-series dataset.

Five models such as Linear Regression, Random Forest, XGBoost, SVM, and ARIMA were evaluated using RMSE, MAE, R², and inference time. Linear Regression delivered the highest accuracy (R² = 0.817) and the fastest inference speed, outperforming more complex models. The pipeline includes KNN-based imputation, sliding-window feature engineering, and model comparison.

While effective under typical pollution conditions, all models struggled with extreme events due to limited representation in the training data. This lightweight forecasting approach offers a practical tool for pollution-prone cities with limited computational infrastructure. Future work includes real-time data integration, deep learning models, and multi-city deployment.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Jameel Syed, Muslim UNSPECIFIED
Subjects:	G Geography. Anthropology. Recreation > GE Environmental Sciences Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning G Geography. Anthropology. Recreation > GE Environmental Sciences > Earth sciences > Atmospheric science > Meteorology
Divisions:	School of Computing > Master of Science in Artificial Intelligence for Business
Depositing User:	Ciara O'Brien
Date Deposited:	24 Jun 2026 11:49
Last Modified:	24 Jun 2026 11:49
URI:	https://norma.ncirl.ie/id/eprint/9405

Actions (login required)

View Item