Comparative Analysis of Multi-Horizon Deep Learning for Water Quality Forecasting in Shrimp Ponds
Proceedings of Engineering and Technology Innovation · 27.08.2026 · 34 · 63-76
DOI: 10.46604/peti.2026.16540 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of Engineering and Technology Innovation, online erschienen am 27.08.2026. Verfasst von Santi und 2 weiteren Autor:innen. Die Arbeit nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Accurate multi-step water-quality prediction is essential for proactive and sustainable aquaculture management. Existing studies often evaluate only limited deep learning (DL) architectures or forecasting horizons, making fair comparisons difficult. This study systematically compares seven DL architectures for forecasting pH, temperature, salinity, and dissolved oxygen (DO) using 6,741 records collected from shrimp ponds at seven Indonesian locations (2017–2025). All models are evaluated using the same preprocessing, optimization, and evaluation settings for forecasting horizons of 5, 10, 20 and 30 steps ahead. Prediction performance consistently decreases as the forecasting horizon increases. Bidirectional long short-term memory (BiLSTM) achieves the lowest root mean square error (RMSE) (2.18) and mean absolute error (MAE) (0.82) at horizon 5 (H5), although repeated experiments indicate that the observed differences are not statistically significant. Performance varies across longer forecasting horizons, with gated recurrent unit (GRU) achieving the lowest RMSE at H20 (2.69).
Zitieren
APA
Santi, Fariza, A., & Gunawan, A. I. (2026). Comparative Analysis of Multi-Horizon Deep Learning for Water Quality Forecasting in Shrimp Ponds. Proceedings of Engineering and Technology Innovation, 34, 63-76. https://doi.org/10.46604/peti.2026.16540
Vancouver
Santi, Fariza A, Gunawan AI. Comparative Analysis of Multi-Horizon Deep Learning for Water Quality Forecasting in Shrimp Ponds. Proceedings of Engineering and Technology Innovation. 2026;34:63-76. doi:10.46604/peti.2026.16540
BibTeX
@article{Santi2026,
author = {Santi and Fariza, Arna and Gunawan, Agus Indra},
title = {Comparative Analysis of Multi-Horizon Deep Learning for Water Quality Forecasting in Shrimp Ponds},
journal = {Proceedings of Engineering and Technology Innovation},
year = {2026},
volume = {34},
pages = {63--76},
doi = {10.46604/peti.2026.16540}
}
EUVIMED