Pajkovski, Darko and Rendevski, Nikola and Ristevski, Blagoj (2025) Next-day Air Pollution Forecasting based on Environmental Factors and Neural Network Models. In: 60th International Scientific Conference on Information, Communication and Energy Systems and Technologies, June 26–28, Ohrid, North Macedonia.
![[thumbnail of ICEST2025-5367.pdf]](https://eprints.uklo.edu.mk/style/images/fileicons/text.png) |
Text
ICEST2025-5367.pdf Restricted to Repository staff only Download (418kB) |
Abstract
Urban regions, with their high population densities,
contend with pervasive air pollution stemming from various
sources
like
vehicular
emissions,
industrial
facilities,
infrastructure, and waste disposal. As a developing country,
Macedonia grapples with significant air quality challenges due to
these diverse pollution contributors. The problem is highly
present in the capital city, Skopje, as well as in other bigger cities
like Bitola, Tetovo and Kumanovo. In this work, we use Neural
Network models to predict the level of each particle
concentration (PM10, PM2.5, O3, CO, NO2, SO2). We are taking
time series quality measurement data from Bitola-2 station and
the meteorological conditions such as temperature, pressure, and
wind. We compare different neural network models'
performances to an Auto-Regressive Integrated Moving Average
(ARIMA) model. The results indicate that the suggested models
consistently surpass the baseline model in performance,
demonstrating their efficacy for accurately predicting air
pollution levels.
| Item Type: | Conference or Workshop Item (Poster) |
|---|---|
| Subjects: | Scientific Fields (Frascati) > Natural sciences > Computer and information sciences |
| Divisions: | Faculty of Information and Communication Technologies |
| Depositing User: | MSc Darko Pajkovski |
| Date Deposited: | 03 Sep 2025 13:53 |
| Last Modified: | 03 Sep 2025 13:53 |
| URI: | https://eprints.uklo.edu.mk/id/eprint/11044 |
Actions (login required)
 |
View Item |