LONG-TERM AIR QUALITY (PM2.5) PREDICTION USING DEEP LEARNING-BASED TEMPORAL CONVOLUTIONAL NETWORKS
Mühendislik Bilimleri ve Tasarım Dergisi · 30.06.2026 · 14 (2) · 416-430
DOI: 10.21923/jesd.1902759 ↗Kurz gefasst
Fachartikel in der Zeitschrift Mühendislik Bilimleri ve Tasarım Dergisi, online erschienen am 30.06.2026. Verfasst von Vakkas Doğan und 2 weiteren Autor:innen. Die Arbeit nennt 33 Literaturangaben.
Abstract (Originalsprache)
Among various sources of pollution, particulate matter with a diameter of 2.5 micrometers or less (PM2.5) is the chief environmental problem worldwide. Collectively, the details of this pollutant PM2.5, its sources and effects measured from various studies, the pollution levels in various cities, the methods used to measure and predict the concentration, the effect of climate change and the possible solution, have been discussed in this review. Pollution levels are rising globally at alarming rates with detrimental effects on the environment as well as human health. However, air quality data represents a time series that is inherently challenging to forecast given its nonlinear nature, noisy data, and long-term dependencies in time. In this work, a Temporal Convolutional Network (TCN)-based deep learning model has been proposed for PM2.5 forecasting, and its performance benchmarked against RNN, LSTM, GRU, and Attention mechanism-based Bi-LSTM models. The study was based on experiments using the hourly dataset of the air quality station in Beijing Aotizhongxin from 2013 to 2017.
Zitieren
APA
Doğan, V., Erten, M. Y., & Aydilek, H. (2026). LONG-TERM AIR QUALITY (PM2.5) PREDICTION USING DEEP LEARNING-BASED TEMPORAL CONVOLUTIONAL NETWORKS. Mühendislik Bilimleri ve Tasarım Dergisi, 14(2), 416-430. https://doi.org/10.21923/jesd.1902759
Vancouver
Doğan V, Erten MY, Aydilek H. LONG-TERM AIR QUALITY (PM2.5) PREDICTION USING DEEP LEARNING-BASED TEMPORAL CONVOLUTIONAL NETWORKS. Mühendislik Bilimleri ve Tasarım Dergisi. 2026;14(2):416-430. doi:10.21923/jesd.1902759
BibTeX
@article{Doan2026,
author = {Doğan, Vakkas and Erten, Mustafa Yasin and Aydilek, Hüseyin},
title = {LONG-TERM AIR QUALITY (PM2.5) PREDICTION USING DEEP LEARNING-BASED TEMPORAL CONVOLUTIONAL NETWORKS},
journal = {Mühendislik Bilimleri ve Tasarım Dergisi},
year = {2026},
volume = {14},
number = {2},
pages = {416--430},
doi = {10.21923/jesd.1902759}
}
EUVIMED