Machine learning may accelerate the recognition and control of microplastic pollution: Future prospects

Area of Science:

• Environmental Science
• Data Science
• Ecotoxicology

Background:

• Microplastics (MPs) are pervasive pollutants across marine, freshwater, terrestrial, and atmospheric environments.
• Evaluating the ecological risks of MPs is complex due to data heterogeneity and a lack of standardized protocols.
• Machine learning (ML) offers a powerful approach to analyze big data for MP risk assessment.

Purpose of the Study:

• To highlight the potential of machine learning in assessing microplastic ecological risks.
• To identify current challenges in microplastic big data research, including data gaps and protocol inconsistencies.
• To propose future directions for ML applications in microplastic research, emphasizing causality and interpretability.

Main Methods:

• Review of current challenges and opportunities in microplastic big data analysis.
• Exploration of machine learning applications for predicting microplastic distribution and ecological impact.
• Discussion of the need for standardized microplastic collection and testing protocols.

Main Results:

• Machine learning can accelerate the evaluation and control of hazardous microplastics.
• Current microplastic databases are insufficient for big data research, necessitating standardized protocols.
• There is a lack of large-scale predictions and interpretable ML models for microplastic environmental risks.

Conclusions:

• Standardized protocols are crucial for advancing microplastic big data research.
• Developing robust, interpretable machine learning models is essential for understanding and mitigating microplastic risks.
• Referencing methods from other particle pollutants can accelerate future microplastic studies.