Scholarlink

<p>To improve water quality prediction, the current investigation examined the effectiveness of combining one optimization algorithm, the Artificial Protozoa Optimization (APO), with an established machine learning (ML) model. The basic models are Decision Tree Classification (DTC) and Extreme Learning Machine (ELM). This strategic integration aimed to increase the prediction performance of the current models by utilizing the optimization power of APO. Based on the evaluation via evaluation metrics including Accuracy, Precision, Recall, and F1-Score, the models’ performance are determined. In the testing phase, the ELAP model achieves the highest Accuracy of 0.9823, Precision of 0.9826, Recall of 0.9823, and F1-score of 0.9824, outperforming all compared models. Under varying water conditions, ELAP maintains superior performance, achieving a maximum Precision of 0.9917 under neutral pH, Recall of 0.9908 in acidic environments, and an F1-score of 0.9896 in acidic conditions.</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 parameter prediction accuracy using a hybrid machine learning and optimization approach

  • 期刊论文

  • 发表日期：

    2025年9月10日

查看全文
  • Minghua Wang

摘要

To improve water quality prediction, the current investigation examined the effectiveness of combining one optimization algorithm, the Artificial Protozoa Optimization (APO), with an established machine learning (ML) model. The basic models are Decision Tree Classification (DTC) and Extreme Learning Machine (ELM). This strategic integration aimed to increase the prediction performance of the current models by utilizing the optimization power of APO. Based on the evaluation via evaluation metrics including Accuracy, Precision, Recall, and F1-Score, the models’ performance are determined. In the testing phase, the ELAP model achieves the highest Accuracy of 0.9823, Precision of 0.9826, Recall of 0.9823, and F1-score of 0.9824, outperforming all compared models. Under varying water conditions, ELAP maintains superior performance, achieving a maximum Precision of 0.9917 under neutral pH, Recall of 0.9908 in acidic environments, and an F1-score of 0.9896 in acidic conditions.