Scholarlink

Access to safe drinking water is the biggest challenge faced by India’s hilly region Ladakh. Ladakh’s Leh district particularly suffers problems of water scarcity, and non-potable water because of the tourism boom and increasing urbanization. This underscores the critical need for effective water harvesting techniques and stringent water quality maintenance. This study examines water quality parameters: pH, temperature, potassium, calcium, chloride, magnesium, total hardness, and alkalinity for Total Dissolved Solids prediction. The proposed framework utilizes interpolation techniques to enrich the dataset and investigates correlations between Total Dissolved Solid and other water quality features. The machine learning models Artificial Neural Networks and K-Nearest Neighbours are deployed to predict the values. The KNN model achieved an impressive \(R^2\) value of 0.9889 with lower errors (MAE of 0.0289 and MSE of 0.0016), significantly outperforming the ANN model, which showed an \(R^2\) value of 0.8713 with MAE of 0.1160 and MSE of 0.0194. These findings indicate that the KNN model is more aptly suited for analyzing water quality in the Leh district, supporting the deployment of this model in further studies to ensure the sustainability of water resources. Strong correlations were observed among parameters such as total hardness, calcium, sodium, chloride, and alkalinity, whereas pH and longitude displayed weaker correlations.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning Techniques on Sustainable Water Management in Ladakh: Performance Analysis of KNN and ANN in Predicting TDS Values

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Nitesh Kumar Sahu,
  • Mridu Sahu,
  • Shivangi Diwan,
  • D. C. Jhariya,
  • Mayank Srivastav,
  • Chandan Singh

摘要

Access to safe drinking water is the biggest challenge faced by India’s hilly region Ladakh. Ladakh’s Leh district particularly suffers problems of water scarcity, and non-potable water because of the tourism boom and increasing urbanization. This underscores the critical need for effective water harvesting techniques and stringent water quality maintenance. This study examines water quality parameters: pH, temperature, potassium, calcium, chloride, magnesium, total hardness, and alkalinity for Total Dissolved Solids prediction. The proposed framework utilizes interpolation techniques to enrich the dataset and investigates correlations between Total Dissolved Solid and other water quality features. The machine learning models Artificial Neural Networks and K-Nearest Neighbours are deployed to predict the values. The KNN model achieved an impressive \(R^2\) value of 0.9889 with lower errors (MAE of 0.0289 and MSE of 0.0016), significantly outperforming the ANN model, which showed an \(R^2\) value of 0.8713 with MAE of 0.1160 and MSE of 0.0194. These findings indicate that the KNN model is more aptly suited for analyzing water quality in the Leh district, supporting the deployment of this model in further studies to ensure the sustainability of water resources. Strong correlations were observed among parameters such as total hardness, calcium, sodium, chloride, and alkalinity, whereas pH and longitude displayed weaker correlations.