A Hybrid Deep Framework for Accurate Water Potability Prediction Using Physicochemical Parameters

Water quality is a critical factor in public health and sustainable water resource management. Rapid industrialization, agricultural expansion, and urban development have significantly increased pressure on water resources, leading to rising contamination levels in many regions. Traditional laboratory-based water quality assessment methods, although accurate, are time-consuming, costly, and difficult to scale for continuous monitoring. Moreover, the nonlinear and multivariate nature of physicochemical water parameters complicates reliable analysis. To address these challenges, this study proposes a hybrid deep learning framework combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks for accurate water potability prediction. The CNN component extracts local feature interactions, while the LSTM captures complex nonlinear dependencies among water quality indicators. Experimental results demonstrate that the proposed model improves Recall and F۱-score while reducing overfitting, achieving an overall accuracy of ۰.۸۶. The findings indicate that the proposed framework can support intelligent water quality monitoring and enhance decision-making in sustainable water resource management.

کلیدواژه ها:

Water Quality ، CNN ، LSTM ، Deep Learning ، Hybrid Deep Framework