Assessment of Water Hydrochemical Parameters Using Machine Learning Tools

Ivan Malashin (), Vladimir Nelyub, Aleksei Borodulin, Andrei Gantimurov and Vadim Tynchenko ()
Additional contact information
Ivan Malashin: Artificial Intelligence Technology Scientific and Education Center, Bauman Moscow State Technical University, 105005 Moscow, Russia
Vladimir Nelyub: Artificial Intelligence Technology Scientific and Education Center, Bauman Moscow State Technical University, 105005 Moscow, Russia
Aleksei Borodulin: Artificial Intelligence Technology Scientific and Education Center, Bauman Moscow State Technical University, 105005 Moscow, Russia
Andrei Gantimurov: Artificial Intelligence Technology Scientific and Education Center, Bauman Moscow State Technical University, 105005 Moscow, Russia
Vadim Tynchenko: Artificial Intelligence Technology Scientific and Education Center, Bauman Moscow State Technical University, 105005 Moscow, Russia

Sustainability, 2025, vol. 17, issue 2, 1-21

Abstract: Access to clean water is a fundamental human need, yet millions of people worldwide still lack access to safe drinking water. Traditional water quality assessments, though reliable, are typically time-consuming and resource-intensive. This study investigates the application of machine learning (ML) techniques for analyzing river water quality in the Barnaul area, located on the Ob River in the Altai Krai. The research particularly highlights the use of the Water Quality Index (WQI) as a key factor in feature engineering. WQI, calculated using the Horton model, integrates nine hydrochemical parameters: pH, hardness, solids, chloramines, sulfate, conductivity, organic carbon, trihalomethanes, and turbidity. The primary objective was to demonstrate the contribution of WQI in enhancing predictive performance for water quality analysis. A dataset of 2465 records was analyzed, with missing values for parameters (pH, sulfate, and trihalomethanes) addressed using predictive imputation via neural network (NN) architectures optimized with genetic algorithms (GAs). Models trained without WQI achieved moderate predictive accuracy, but incorporating WQI as a feature dramatically improved performance across all tasks. For the trihalomethanes model, the R 2 score increased from 0.68 (without WQI) to 0.86 (with WQI). Similarly, for pH, the R 2 improved from 0.35 to 0.74, and for sulfate, from 0.27 to 0.69 after including WQI in the feature set.

Keywords: water quality index; machine learning; predictive modeling (search for similar items in EconPapers)
JEL-codes: O13 Q Q0 Q2 Q3 Q5 Q56 (search for similar items in EconPapers)
Date: 2025
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2071-1050/17/2/497/pdf (application/pdf)
https://www.mdpi.com/2071-1050/17/2/497/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jsusta:v:17:y:2025:i:2:p:497-:d:1564080

Access Statistics for this article

Sustainability is currently edited by Ms. Alexandra Wu

More articles in Sustainability from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().