Pregled bibliografske jedinice broj: 1252646
Short-term prediction of sea levels by supervised learning models
Short-term prediction of sea levels by supervised learning models // Proceedings of the 3rd IAHR Young Professionals Congress
online: International Association for Hydro-Environment Engineering and Research -IAHR, 2022. str. 156-157 (predavanje, međunarodna recenzija, prošireni sažetak, znanstveni)
CROSBI ID: 1252646 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Short-term prediction of sea levels by supervised
learning models
Autori
Berbić, Jadran ; Šreng, Željko
Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, prošireni sažetak, znanstveni
Izvornik
Proceedings of the 3rd IAHR Young Professionals Congress
/ - : International Association for Hydro-Environment Engineering and Research -IAHR, 2022, 156-157
ISBN
978-90-832612-8-7
Skup
3rd IAHR Young Professionals Congress
Mjesto i datum
Online, 28.11.2022. - 02.12.2022
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
supervised learning ; successive halving ; sea levels ; short-term prediction
Sažetak
In the work six (6) supervised learning algorithms were used for prediction of sea levels at the mareographic station Split (Croatia). Model’s parameters were optimized by successive halving for prediction of 1, 2, 3, 6, 12 and 24 hours ahead. Highest coefficients of determination were produced by gradient boosting regressor and histogram gradient boosting regressor. Considering the duration of SH procedure, advantage could be given to artificial neural network due to the slightly smaller accuracies but significantly less duration.
Izvorni jezik
Engleski
Znanstvena područja
Građevinarstvo
POVEZANOST RADA
Ustanove:
Građevinski i arhitektonski fakultet Osijek
