Artificial intelligence-based optimization and modeling of cadmium reduction via ultraviolet-assisted malathion/sulfite reaction mechanisms
This study focuses on optimizing cadmium reduction through the UV/malathion/sulfite reaction, leveraging the power of Artificial Intelligence (AI) models, specifically Gradient Boosting Regression (GBR), Support Vector Regression (SVR), and Genetic Algorithm (GA). These models were used to optimize...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-07-01
|
| Series: | Results in Chemistry |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2211715625004722 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study focuses on optimizing cadmium reduction through the UV/malathion/sulfite reaction, leveraging the power of Artificial Intelligence (AI) models, specifically Gradient Boosting Regression (GBR), Support Vector Regression (SVR), and Genetic Algorithm (GA). These models were used to optimize the reduction process and attain the requisite conditions from the cadmium hexavalent (Cd (VI)) to trivalent (Cd (III)). Regarding the training set evaluation, high accuracy was observed with minimal error and R2 of 0.996. However, the GBR model tended to overlearn the test data as evidenced by a substantial decrease in its general performance. In contrast, the SVR model showed better generalization to the test data, with lower error metrics and a consistent R2 of around 0.81 across both datasets. The GA model refined the input parameters, enhanced the other phases of the removal of cadmium, and further optimized the overall process. Feature importance analysis revealed that the most influential factors in both models were related to the concentrations of malathion, and sulfite, with malathion (X4) emerging as the most critical parameter. The results highlight the potential of AI models in optimizing environmental processes, with SVR showing better generalization capabilities and GBR providing insights into the relationships between input parameters. This study offers valuable insights into using AI models for optimizing pollutant removal processes, demonstrating their effectiveness in environmental applications such as cadmium reduction. |
|---|---|
| ISSN: | 2211-7156 |