Scholarlink

<p>This study introduces an innovative approach to water quality monitoring by integrating Explainable AI (XAI) with Wasserstein Generative Adversarial Networks (WGAN) for data augmentation. Traditional water quality monitoring methods often struggle with limited and imbalanced datasets, leading to challenges in accurate model training and prediction. To address this, WGAN is employed to generate realistic synthetic data, enriching the dataset and improving model robustness. Concurrently, XAI techniques are applied to ensure that the model’s decision-making process is transparent and interpretable, enabling stakeholders to understand how specific features impact water quality predictions. The proposed approach is validated through a case study at the PIL site, where it demonstrates superior performance in predicting water quality metrics compared to conventional methods. This combination of WGAN-based augmentation and XAI offers a powerful tool for enhancing the accuracy, reliability, and trustworthiness of water quality monitoring systems, providing actionable insights for environmental management and decision-making.</p>

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

Enhancing Water Quality Monitoring with Explainable AI and WGAN-Based Data Augmentation

  • 期刊论文

  • 发表日期：

    2025年2月14日

查看全文
  • Shofia Priyadharshini D.,
  • G. P. Ramesh

摘要

This study introduces an innovative approach to water quality monitoring by integrating Explainable AI (XAI) with Wasserstein Generative Adversarial Networks (WGAN) for data augmentation. Traditional water quality monitoring methods often struggle with limited and imbalanced datasets, leading to challenges in accurate model training and prediction. To address this, WGAN is employed to generate realistic synthetic data, enriching the dataset and improving model robustness. Concurrently, XAI techniques are applied to ensure that the model’s decision-making process is transparent and interpretable, enabling stakeholders to understand how specific features impact water quality predictions. The proposed approach is validated through a case study at the PIL site, where it demonstrates superior performance in predicting water quality metrics compared to conventional methods. This combination of WGAN-based augmentation and XAI offers a powerful tool for enhancing the accuracy, reliability, and trustworthiness of water quality monitoring systems, providing actionable insights for environmental management and decision-making.