A Literature Review of Machine Learning Algorithm Performance in Water Quality Parameter Prediction

Water quality is essential for environmental sustainability and public health, yet conventional laboratory-based monitoring is costly and slow. Machine learning (ML) has been proposed as an alternative, but algorithm performance is inconsistent across studies and seldom compared on identical target parameters. This study synthesizes ML algorithm use and performance in predicting water quality parameters. A structured narrative literature review searched Scopus for English-language publications from 2015 to 2026 using keywords combining machine learning, deep learning, or artificial intelligence with water quality and prediction, forecasting, or classification. Titles, abstracts, and full texts were screened against predefined inclusion and exclusion criteria; 37 publications were analyzed. Algorithms, target parameters, validation strategies, and metrics (R², RMSE, MAE) were extracted and synthesized narratively, with performance grouped by target parameter, not compared across targets. Random Forest, Support Vector Machine, Decision Tree, K-Nearest Neighbors, neural networks, and boosting algorithms (XGBoost, AdaBoost, CatBoost, LightGBM) were most frequently used, while the Water Quality Index (WQI) was the most frequently predicted target, followed by dissolved oxygen (DO), pH, and biochemical oxygen demand (BOD). Ensemble and boosting models generally performed strongly; for example, XGBoost reached R² = 0.93 and RMSE = 0.047 for the Irrigation Water Quality Index, and an attention-based LSTM reached R² = 0.985 for daily DO. However, no algorithm was universally superior, and external validation remained rare. Algorithm selection should consider data characteristics, target parameter, evaluation metrics, and interpretability.

Authors

Institutions

Publication Details

Journal
TRANSPUBLIKA INTERNATIONAL RESEARCH IN EXACT SCIENCES
Published
2026-11-19
DOI
https://doi.org/10.55047/tires.v5i3.2476
Primary Topic
Hydrological Forecasting Using AI
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

A Literature Review of Machine Learning Algorithm Performance in Water Quality Parameter Prediction

Asep Saefurohman
TRANSPUBLIKA INTERNATIONAL RESEARCH IN EXACT SCIENCES
Hydrological Forecasting Using AI
article

A Literature Review of Machine Learning Algorithm Performance in Water Quality Parameter Prediction

Asep Saefurohman
article en

Abstract

Water quality is essential for environmental sustainability and public health, yet conventional laboratory-based monitoring is costly and slow. Machine learning (ML) has been proposed as an alternative, but algorithm performance is inconsistent across studies and seldom compared on identical target parameters. This study synthesizes ML algorithm use and performance in predicting water quality parameters. A structured narrative literature review searched Scopus for English-language publications from 2015 to 2026 using keywords combining machine learning, deep learning, or artificial intelligence with water quality and prediction, forecasting, or classification. Titles, abstracts, and full texts were screened against predefined inclusion and exclusion criteria; 37 publications were analyzed. Algorithms, target parameters, validation strategies, and metrics (R², RMSE, MAE) were extracted and synthesized narratively, with performance grouped by target parameter, not compared across targets. Random Forest, Support Vector Machine, Decision Tree, K-Nearest Neighbors, neural networks, and boosting algorithms (XGBoost, AdaBoost, CatBoost, LightGBM) were most frequently used, while the Water Quality Index (WQI) was the most frequently predicted target, followed by dissolved oxygen (DO), pH, and biochemical oxygen demand (BOD). Ensemble and boosting models generally performed strongly; for example, XGBoost reached R² = 0.93 and RMSE = 0.047 for the Irrigation Water Quality Index, and an attention-based LSTM reached R² = 0.985 for daily DO. However, no algorithm was universally superior, and external validation remained rare. Algorithm selection should consider data characteristics, target parameter, evaluation metrics, and interpretability.

TRANSPUBLIKA INTERNATIONAL RESEARCH IN EXACT SCIENCESVol. 5(3)
Universitas Islam Negeri Sultan Maulana Hasanuddin Banten (ID)
Hydrological Forecasting Using AI
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon