**4. Conclusions**

Implementing machine learning models enables the analysis of water turbidity behavior and facilitates decision-making in treatment plants with limited instrumentation. The results indicate no linear relationship between the database parameters and turbidity, which led to the evaluation of non-linear models that better estimate water turbidity. Although these models show slightly lower performance when reducing the number of predictors, they allow for the appreciation of climatic influence and can be applied more efficiently in treatment plants with limited resources. In addition, by setting a limit on turbidity to exclude outlier values, the performance of the models was improved. The evaluation of models that consider outlier values and achieve better performance could be the subject of future research.

**Author Contributions:** Conceptualization, J.F.C. and J.C.C.; methodology, C.F., V.F.A. and D.G.S.; software, V.F.A.; formal analysis, V.F.A.; investigation, V.F.A. and D.G.S.; resources, J.F.C.; writing— original draft preparation, V.F.A. and D.G.S.; writing—review and editing, V.F.A., D.G.S. and C.F.; supervision, C.F.; project administration, V.F.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data Availability Statements are available in section "water treatment plant" at https://www.kaggle.com/datasets/jvanessafernandez/water-treatment-plant, accessed on 2 April 2023.

**Acknowledgments:** We would like to express our gratitude to the Aquarisc project for their valuable collaboration in providing data for the development of this project.

**Conflicts of Interest:** The authors declare no conflict of interest.
