Smart Multisensor-Based Machine Learning for Multiclass Freshwater Fishpond Water Quality Degradation Classification
Received: 29 January 2026 | Revised: 27 February 2026, 10 March 2026, and 13 March 2026 | Accepted: 15 March 2026 | Online: 12 April 2026
Corresponding author: Giva Andriana Mutiara
Abstract
Water quality degradation in freshwater fishponds results from interactions among physicochemical parameters, limiting the effectiveness of single-parameter threshold monitoring. This study presents a multisensor-based Machine Learning (ML) framework for multiclass classification of water quality degradation in a monitored earthen freshwater pond. A dataset comprising 2153 data points was labeled into four degradation levels (Normal, Caution, Warning, Severe) using a predefined rule-based multi-parameter scoring protocol derived from aquaculture operational standards, ensuring independence from model training. Seven physicochemical parameters—Oxidation–Reduction Potential (ORP), Electrical Conductivity (EC), Total Dissolved Solids (TDS), turbidity, temperature, pH, and Dissolved Oxygen (DO)—were analyzed. Four baseline classifiers: Random Forest (RF), XGBoost, Support Vector Machine (SVM) with RBF kernel, and distance-weighted K-Nearest Neighbors (KNN) were evaluated using stratified 5-fold cross-validation, with performance reported as mean ± standard deviation. RF achieved the highest performance, obtaining an accuracy of 0.952 ± 0.013 and a macro-F1 score of 0.955 ± 0.012. To address multicollinearity, Pearson correlation analysis and targeted EC–TDS feature ablation were conducted. The results indicate minimal performance variation when either EC or TDS is removed individually, with noticeable degradation only when both are excluded. SHapley Additive Explanations (SHAP)-based explainability and Principal Component Analysis (PCA) further identify turbidity and EC as dominant degradation indicators. The first two main components capture approximately 58–60% of variance, indicating a multidimensional degradation structure rather than complete separability in two-dimensional space. The proposed framework offers a structured and reproducible engineering-oriented approach for assessing multiclass freshwater fishpond degradation under moderate class imbalance.
Keywords:
water quality degradation, freshwater fishpond monitoring, multisensor systems, machine learning classification, random forestDownloads
References
R. H. Bosma and M. C. J. Verdegem, "Sustainable Aquaculture in Ponds: Principles, Practices and Limits," Livestock Science, vol. 139, no. 1–2, pp. 58–68, Jul. 2011.
W. Wang, Y. Xu, S. Zhang, Y. Xu, and X. Xue, "Bibliometric Insights into Pollution Research: Trends, Geographic Disparities, and Emerging Environmental Challenges," Environmental Research Communications, vol. 7, no. 7, Jul. 2025, Art. no. 075026.
N. Turlybek et al., "Sustainable Aquaculture Systems and Their Impact on Fish Nutritional Quality," Fishes, vol. 10, no. 5, May 2025, Art. no. 206.
B. Bojarski et al., "The Influence of Fish Ponds on Fish Assemblages of Adjacent Watercourses," Polish Journal of Environmental Studies, vol. 31, no. 1, pp. 609–617, Jan. 2022.
M. Al Tulus and D. Maulianawati, "Assessment of Water Quality Parameters in Penaeus monodon Culture Ponds: Implications for Sustainable Shrimp Aquaculture," Aquatic Life Sciences, vol. 2, no. 2, pp. 64–69, Dec. 2025.
Y. Avnimelech, "Evaluating the Active Redox and Organic Fractions in Pond Bottom Soils: EOM, Easily Oxidized Material," Aquaculture, vol. 233, no. 1–4, pp. 283–292, Apr. 2004.
Md. M. Hemal et al., "An Integrated Smart Pond Water Quality Monitoring and Fish Farming Recommendation Aquabot System," Sensors, vol. 24, no. 11, Jun. 2024, Art. no. 3682.
A. N. Laghari, G. D. Walasai, A. R. Jatoi, F. A. Shaikh, and Z. A. Siyal, "Performance Analysis of Water Filtration Units for Reduction of pH, Turbidity, Solids and Electricity Conductivity," Engineering, Technology & Applied Science Research, vol. 8, no. 4, pp. 3209–3212, Aug. 2018.
Safety and Quality of Water Used in the Production and Processing of Fish and Fishery Products. Rome, Italy: Food and Agriculture Organization of the United Nations World Health Organization, 2023.
Hamzah et al., "Analysis of the Effectiveness of Sand Filters for Raw Water Treatment in Brackish Water Aquaculture Environments," Engineering, Technology & Applied Science Research, vol. 15, no. 6, pp. 29448–29456, Dec. 2025.
Z. Bohdan, P. Oleksandr, S. Volodymyr, and S. Viktor, "Machine Learning-Based Environmental Monitoring and Analysis System," in Workshop on Cybersecurity Providing in Information and Telecommunication Systems, Kyiv, Ukraine, Feb. 2025, pp. 183–203.
S. Ramya, S. Srinath, and P. Tuppad, "Comprehensive Analysis of Multiple Classifiers for Enhanced River Water Quality Monitoring with Explainable AI," Case Studies in Chemical and Environmental Engineering, vol. 10, Art. no. 100822, Dec. 2024.
K. Sethy, "Predicting Water Quality Using Ensemble Machine Learning Models and Remote Sensing Data," International Journal of Environmental Sciences, pp. 166–175, Jul. 2025.
E. Alotaibi and N. Nassif, "Artificial Intelligence in Environmental Monitoring: In-Depth Analysis," Discover Artificial Intelligence, vol. 4, no. 1, Nov. 2024, Art. no. 84.
H. A. Salman, A. Kalakech, and A. Steiti, "Random Forest Algorithm Overview," Babylonian Journal of Machine Learning, vol. 2024, pp. 69–79, Jun. 2024.
S. Hakkal and A. A. Lahcen, "XGBoost to Enhance Learner Performance Prediction," Computers and Education: Artificial Intelligence, vol. 7, Dec. 2024, Art. no. 100254.
Md. M. Islam, M. A. Kashem, S. A. Alyami, and M. A. Moni, "Monitoring Water Quality Metrics of Ponds with IoT Sensors and Machine Learning to Predict Fish Species Survival," Microprocessors and Microsystems, vol. 102, Oct. 2023, Art. no. 104930.
Md. A. A. M. Hridoy et al., "Advanced Machine Learning Models for Accurate Water Quality Classification and WQI Prediction: Implications for Aquatic Disease Risk Management," Science of the Total Environment, vol. 1008, Dec. 2025, Art. no. 180965.
I. A. Siahaan, G. A. Mutiara, and M. I. Sani, "A Low-Cost Water Quality Monitoring Based on Photodiode and LDR," in 2021 IEEE Asia Pacific Conference on Wireless and Mobile, Apr. 2021, pp. 141–146.
A. Pala, A. Oleynik, and N. O. Handegard, "Feature-Based Cluster Selection Framework for Binary Classification on Imbalanced Acoustic Data," Fisheries Research, vol. 292, Dec. 2025, Art. no. 107598.
V. J. Almero, R. Concepcion, M. Rosales, R. R. Vicerra, A. Bandala, and E. Dadios, "An Aquaculture-Based Binary Classifier for Fish Detection Using Multilayer Artificial Neural Network," in 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management, Laoag, Philippines, Nov. 2019, pp. 1–5.
A. T. Demetillo and E. B. Taboada, "Real-Time Water Quality Monitoring for Small Aquatic Area Using Unmanned Surface Vehicle," Engineering, Technology & Applied Science Research, vol. 9, no. 2, pp. 3959–3964, Apr. 2019.
A. A. Alsakran et al., "Enhancing Interpretability and Explainability for Fish Farmers: Decision Tree Approximation of DDPG for RAS Control," Aquaculture International, vol. 33, no. 6, Nov. 2025, Art. no. 623.
P. Saravanan et al., "Comprehensive Review on Toxic Heavy Metals in the Aquatic System: Sources, Identification, Treatment Strategies, and Health Risk Assessment," Environmental Research, vol. 258, Oct. 2024, Art. no. 119440.
S. M. Naim, P. Das, J.-J. Tiang, and A.-A. Nahid, "Aquaculture Water Quality Classification Using XGBoost Classifier Model Optimized by the Honey Badger Algorithm with SHAP and DiCE-Based Explanations," Water, vol. 17, no. 20, Oct. 2025, Art. no. 2993.
S. S. Habib et al., "Assessment of Heavy Metal Levels in Polyculture Fish Farms and Their Aquatic Ecosystems: An Integrative Study Addressing Environmental and Human Health Risks Associated with Dam Water Usage," Environmental Geochemistry and Health, vol. 46, no. 8, Aug. 2024, Art. no. 267.
G. A. Mutiara, M. R. Alfarisi, and L. Meisaroh, "Freshwater Fishpond Multisensor Water Quality Dataset for Multiclass Degradation Classification (2153 Samples)." Zenodo, Mar. 2026, [Online]. Available: https://zenodo.org/records/19210095.
Downloads
How to Cite
License
Copyright (c) 2026 Giva Andriana Mutiara, Muhammad Rizqy Alfarisi, Lisda Meisaroh
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain the copyright and grant the journal the right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) after its publication in ETASR with an acknowledgement of its initial publication in this journal.
