Advancing Water Quality Research: K-Nearest Neighbor Coupled with the Improved Grey Wolf Optimizer Algorithm Model Unveils New Possibilities for Dry Residue Prediction

Abstract

Monitoring stations have been established to combat water pollution, improve the ecosystem, promote human health, and facilitate drinking water production. However, continuous and extensive monitoring of water is costly and time-consuming, resulting in limited datasets and hindering water management research. This study focuses on developing an optimized K-nearest neighbor (KNN) model using the improved grey wolf optimization (I-GWO) algorithm to predict dry residue quantities. The model incorporates 20 physical and chemical parameters derived from a dataset of 400 samples. Cross-validation is employed to assess model performance, optimize parameters, and mitigate the risk of overfitting. Four folds are created, and each fold is optimized using 11 distance metrics and their corresponding weighting functions to determine the best model configuration. Among the evaluated models, the Jaccard distance metric with inverse squared weighting function consistently demonstrates the best performance in terms of statistical errors and coefficients for each fold. By averaging predictions from the models in the four folds, an estimation of the overall model performance is obtained. The resulting model exhibits high efficiency, with remarkably low errors reflected in the values of R, R2, R2ADJ, RMSE, and EPM, which are reported as 0.9979, 0.9958, 0.9956, 41.2639, and 3.1061, respectively. This study reveals a compelling non-linear correlation between physico-chemical water attributes and the content of dry tailings, indicating the ability to accurately predict dry tailing quantities. By employing the proposed methodology to enhance water quality models, it becomes possible to overcome limitations in water quality management and significantly improve the precision of predictions regarding critical water parameters.

Efficiency of Hydrogen Peroxide and Fenton Reagent for Polycyclic Aromatic Hydrocarbon Degradation in Contaminated Soil: Insights from Experimental and Predictive Modeling

“…This exploration extended further, delving into a myriad of basis functions, including constant, linear, Pure Quadratic, and zero, while meticulously fine-tuning the parameters-Kernel Scale [sigmaM, sigmaF], and sigma-of each kernel function to achieve optimal performance. Amidst this intricate optimization process, the assessment of the model's efficacy hinged upon two pivotal criteria: the coefficient of correlation, a measure of the model's ability to capture underlying trends, and the root mean square error (RMSE), a quantification of the model's predictive accuracy [29,33,34].…”

“…Initially, the experimental and predicted values were overlaid on a graph (Figure 8a), enabling a visual comparison to assess their alignment. This graphical representation facilitated the identification of any systematic deviations or patterns in the model's predictions across different phases [33]. Additionally, a second analytical approach was employed, where an error histogram was constructed to visualize the distribution of residuals for each phase (Figure 8b) [28,33].…”

“…This graphical representation facilitated the identification of any systematic deviations or patterns in the model's predictions across different phases [33]. Additionally, a second analytical approach was employed, where an error histogram was constructed to visualize the distribution of residuals for each phase (Figure 8b) [28,33]. This method allowed for a deeper exploration of the discrepancies between predicted and observed values, shedding light on potential outliers or areas of inconsistency.…”

“…The Improved Grey Wolf Optimizer algorithm (IGWO) emerged as a cornerstone in achieving optimization [33,34]. Leveraging its robust optimization capabilities, the algorithm facilitated the attainment of an optimal solution (minimization of C/C 0 ), which was subsequently rigorously validated within the laboratory setting.…”

Comparative Machine Learning Models for Predicting Loan Fructification in a Semi-Urban Area

“…and then assigning the object in question to one of these categories (Tahraoui et al, 2023). Recently KNN has been used to predict breast cancer and its performance in such prediction revealed to be accurate at 100%.…”