.png){kind=icon}
.png){kind=icon}
Return
Improving Water Quality Evaluation Using an Optimized Multi-Output Support Vector Machine
.png){kind=icon}
PRE
AI
DOI：10.1007/s11269-026-04798-7.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Rapid urbanization and industrial emissions pose serious organic and nutrient pollution risks to the Yangtze River, China’s most important strategic water source and ecological corridor. Although water quality evaluation is vital for pollution control, traditional methods are cumbersome and incapable of simultaneous predictions across multiple stations. To address these challenges, this study proposes a coherent and comprehensive framework rather than a simple list of technologies. Firstly, the entropy-weighted water quality index (EWQI) replaces the traditional subjective water quality index, providing an objective and simplified assessment method. Secondly, to overcome the inefficiency of single station model, a multi-output support vector machine (MSVM) was adopted to simultaneously predict the EWQI across multiple stations, and the optimal combination of environmental factors was identified through correlation analysis. Furthermore, to obtain the optimal hyperparameters of MSVM, the particle swarm optimization (PSO) and the grey wolf optimizer (GWO) were respectively embedded into this framework for comparative analysis. The results show that the proposed GWO-MSVM framework exhibits the best performance with an average Root Mean Square Error (RMSE) of 1.331, an average Nash-Sutcliffe efficiency (NSE) of 0.9603 and an average Coefficient of Determination (R²) of 0.9792. Compared with the baseline MSVM, it reduces the RMSE by 34.6%, and improves the NSE by 8.2%, R² value by 3.5% respectively. This study provides an efficient and accurate multi-station prediction framework, significantly reducing the computational cost of regional monitoring, offering robust technological backing for dynamic water quality forecasting and coordinated watershed-level governance.
Keywords:
Water quality
Water resources management
Machine learning
EWQI
MSVM
Journal
IF:
4.7
Papers:
8.1K
Citations:
1.6W
Organization
View more organizations