Scholarlink

<p>Water bodies are getting highly polluted by heavy metals, particularly copper, which is a severe issue of concern for the environment that requires an effective and sustainable method of remediation. Though traditional machine learning models performed well at predicting the percentage removal of contaminants, their main weakness is not accounting for the temporal lag inherent in chemical adsorption and filtration. A hybrid framework of Gated Recurrent Unit and Convolutional Neural Network to predict the percentage removal of copper were used. CNN offers automated feature extraction, and GRU processes them as a sequence using its gating mechanism, retaining previous time-step information in the current prediction. The model is trained with 80% training data and 20% test dataset, and a 5-fold cross-validation strategy is employed to validate the model’s performance, reliability and its generalization capability. The proposed integrated network is compared with several traditional models, including Multivariate Linear Regression (SM-LR), Bayesian Extreme Gradient Boosting (B-XGM), and Kernel Extreme Learning Machine (K-ELM). A statistical validation using Legates’ Modulus (LM), Willmott’s Index (WI) and Nash-Sutcliffe Efficiency presented along with metrics, mean squared error (MSE), mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R² score. The GRU-CNN outperformed other models with the highest R<sup>2</sup> scores of 0.985 and 0.981 on the training and test datasets with lowest MAE of 1.88 and 2.11. It also achieved the highest WI, LM, and NSE scores of 0.996. 0.868 and 0.981 on the test data. Thus, the proposed experimental and simulation approach offers a scalable method for removing copper from wastewater.</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

Evaluating the Performance of Traditional Predictive Models Against Advanced Machine Learning Techniques for the Removal of Copper Using Low Cost Adsorbent

  • 期刊论文

  • 发表日期：

    2026年3月14日

查看全文
  • Nayeemuddin Mohammed,
  • Hiren Mewada,
  • Feroz Shaik

摘要

Water bodies are getting highly polluted by heavy metals, particularly copper, which is a severe issue of concern for the environment that requires an effective and sustainable method of remediation. Though traditional machine learning models performed well at predicting the percentage removal of contaminants, their main weakness is not accounting for the temporal lag inherent in chemical adsorption and filtration. A hybrid framework of Gated Recurrent Unit and Convolutional Neural Network to predict the percentage removal of copper were used. CNN offers automated feature extraction, and GRU processes them as a sequence using its gating mechanism, retaining previous time-step information in the current prediction. The model is trained with 80% training data and 20% test dataset, and a 5-fold cross-validation strategy is employed to validate the model’s performance, reliability and its generalization capability. The proposed integrated network is compared with several traditional models, including Multivariate Linear Regression (SM-LR), Bayesian Extreme Gradient Boosting (B-XGM), and Kernel Extreme Learning Machine (K-ELM). A statistical validation using Legates’ Modulus (LM), Willmott’s Index (WI) and Nash-Sutcliffe Efficiency presented along with metrics, mean squared error (MSE), mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R² score. The GRU-CNN outperformed other models with the highest R2 scores of 0.985 and 0.981 on the training and test datasets with lowest MAE of 1.88 and 2.11. It also achieved the highest WI, LM, and NSE scores of 0.996. 0.868 and 0.981 on the test data. Thus, the proposed experimental and simulation approach offers a scalable method for removing copper from wastewater.