1. Water quality prediction using data-driven models case study: Ardabil plain, Iran.
- Author
-
Hasanpour Kashani, Mahsa, Nikpour, Mohammad Reza, and Jalali, Reza
- Subjects
- *
WATER quality , *SUPPORT vector machines , *REGRESSION analysis , *PLAINS - Abstract
This research investigates the capabilities of four data-driven models, namely artificial neural network, wavelet-based artificial neural network (WANN), gene expression programming, and support vector machine, to predict water quality parameter, i.e., sodium absorption ratio (SAR) of the groundwater of the Ardabil plain, Ardabil Province, Iran. A combination of data sets including electrical conductivity, total dissolved solids, and sodium was considered as input data of the models. The models' performances were compared with the classical multiple linear regression model. The models were calibrated and validated using the measured quality data for the period of year 2005 to 2015. The performance of the models was evaluated using various statistical criteria, scatter, and PDF plots. Results showed that the data-driven models are more capable in simulating the SAR data than the classical multiple linear regression model. The WANN model provided reliable predictions of the SAR values rather than the other models. [ABSTRACT FROM AUTHOR]
- Published
- 2023
- Full Text
- View/download PDF