River Water Salinity Prediction Utilized a Social Learning and Wavelet Mutation Whale Optimization Algorithm

River water demand forecasting can contribute significantly to water resource management. Analyzing river water salinity prediction involves utilizing datasets on water quality, which can be influenced by various factors, including population growth, economic development, and living standards. However, there are challenges such as susceptibility to falling into local optima, dealing with large input spaces and the increased dimensionality of computational processes, especially with initial clustering input values. To address these challenges, salinity prediction utilizes the Adaptive Neuro Fuzzy Inference System (ANFIS) combined with the Social Learning and Wavelet Mutation Whale Optimization Algorithm (SMWOA) to estimate water resource demand. The proposed method integrates the SMWOA mutation strategies to enhance linear probability and improve global search capabilities. Results demonstrate that the proposed method ANFIS-SMWOA, achieves superior performance metrics including the Nash-Sutcliffe Efficiency Coefficient (NSE) of 0.996, Root Mean Square Error (RMSE) of $\mathbf{0. 3 2 9 5}$, Coefficient of Determination $R^{2}$ of $\mathbf{0. 9 4 1}$, Normalized RMSE (NRMSE) of $\mathbf{0. 1 2 8,}$, Modified Index of Agreement (md) and Kling-Gupta efficiency (KGE) of 0.983 on water quality datasets. These outcomes confirm the better detection and classification performance of the proposed method compared to other existing methods, such as Relevance Vector Machine (RVM), Improved Grasshopper Optimization Algorithm (IGOA), Simulated Annealing (SA), and the Mayfly Optimization Algorithm (MOA).