An insight into machine learning models era in simulating soil, water bodies and adsorption heavy metals: Review, challenges and solutions

Area of Science:

• Environmental Science
• Computational Chemistry

Background:

• Computer-aided models for heavy metals (HMs) simulation have advanced significantly.
• Machine learning (ML) models show outstanding progress in HM modeling over the last two decades.

Purpose of the Study:

• To provide an informative overview of the progression of computer-aided HM simulation models.
• To review and discuss methodological and conceptual aspects of HM modeling in contaminated soil, water, and removal from aqueous solutions.

Main Methods:

• Review of diverse ML models and their application in HM simulation.
• Discussion of key modeling aspects including dataset types, parameter selection, tuning, and performance metrics.

Main Results:

• Identified limitations of classical analytical methods compared to ML approaches.
• Highlighted the necessity for exploring new ML model versions and optimizing existing ones.

Conclusions:

• Emphasized the importance of addressing modeling aspects for credibility and consistency in environmental science.
• Outlined future research directions to enhance understanding of ML applications in HM simulation.