We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Efficient Prediction of Microplastic Counts from Mass Measurements
Summary
Researchers evaluated machine learning models including linear regression, kernel ridge regression, and decision trees for predicting microplastic particle counts from aggregate mass measurements, testing on synthetic and experimental datasets. They found that kernel ridge regression performed best, with lower prediction error for larger and more homogeneous samples, and that organic contamination did not substantially reduce predictive accuracy.
Microplastics must be characterized and quantified to assess their impact. This is complicated by the time-consuming and error-prone nature of current quantification procedures. This study evaluates the use of machine learning to estimate the number of microplastic particles on the basis of aggregate particle weight measurements. Synthetic data sets are used to test the performance of linear regression, kernel ridge regression, and decision trees. Kernel ridge regression, which achieves the best performance, is tested on several experimental datasets. The numerical results show that the algorithm is better at predicting the counts of larger and more homogeneous samples and that contamination by organics does not significantly increase error. In mixed samples, prediction error is lower for heavier particles, with an error rate comparable to or better than that of traditional manual counting.
Sign in to start a discussion.
More Papers Like This
Efficient Prediction of Microplastic Counts from Mass Measurements
Scientists developed machine learning models to estimate the number of microplastic particles from aggregate weight measurements, potentially offering a faster and cheaper alternative to manual counting. Efficient quantification methods are critical for large-scale monitoring of microplastic contamination in environmental samples.
Prediction of Settling Velocity of Microplastics by Multiple Machine-Learning Methods
Researchers developed machine learning models to predict the settling velocity of microplastics in water, using particle shape, size, and density as inputs. The models outperformed traditional empirical equations, providing a more accurate tool for modeling microplastic transport and sedimentation.
An Accurate Size-Probability Distribution Method for Converting Microplastic Counts to Mass
Researchers developed a size-probability distribution method to convert microplastic particle counts into mass estimates without requiring detailed morphological measurements for every particle, addressing a key gap in environmental monitoring where mass-based reporting is needed but count-based data is more commonly generated.
Convex Optimization of Environmental Processes
Researchers developed a Kernel Ridge Regression (KRR)-based method enhanced by active learning strategies to predict microplastic particle counts from aggregate weight measurements, reducing the need for labor-intensive manual counting by training on a subset of experimentally quantified samples and using active learning to minimize training data requirements without sacrificing prediction accuracy.
Predicting the toxicity of microplastic particles through machine learning models
Researchers developed machine learning models to predict microplastic particle toxicity from physical and chemical descriptors, addressing the classification challenge posed by the enormous diversity of particle types that cannot be characterized using conventional chemical hazard methods. The models provided accurate toxicity predictions across diverse microplastic types, offering a practical screening tool for the field.