Related Experiment Video
Updated: Jan 19, 2026
13:18
Data Collection on Marine Litter Ingestion in Sea Turtles and Thresholds for Good Environmental Status
Published on: May 18, 2019
12.6K
Rummaging through the bin: Modelling marine litter distribution using Artificial Neural Networks
S Franceschini1, F Mattei1, L D'Andrea1
1Laboratory of Experimental Ecology and Aquaculture, Department of Biology, University of Rome Tor Vergata, via della Ricerca Scientifica snc, 00133 Rome, Italy; CoNISMa, Piazzale Flaminio, 9, 00196 Rome, Italy.
Marine Pollution Bulletin
|September 24, 2019
Summary
Artificial Neural Networks can model marine litter distribution and quantity on the seabed. Machine learning offers a promising approach to assess marine litter issues and identify hotspots in marine environments.
Area of Science:
- Marine Biology
- Environmental Science
- Data Science
Background:
- Marine litter poses significant ecological, social, and economic threats.
- Accurate assessment of marine litter hotspots and accumulation zones remains a challenge.
- Predictive models for seabed marine litter distribution are currently limited.
Purpose of the Study:
- To model the influence of environmental factors on marine litter distribution.
- To estimate the total quantity of marine litter on the seabed in the Central Mediterranean Sea.
- To evaluate the efficacy of Artificial Neural Networks (ANNs) for marine litter assessment.
Main Methods:
- Utilized Artificial Neural Networks (ANNs), including Self-Organing Maps (SOMs) and Multilayer Perceptrons (MLPs).
- Employed environmental descriptors to model the relationship with marine litter density.
- Developed an MLP model for quantifying regional seabed marine litter.
Main Results:
- Self-Organizing Maps identified key environmental descriptors influencing marine litter density.
- The Multilayer Perceptron model demonstrated efficiency in estimating regional seabed litter quantities.
- Results indicate a strong correlation between environmental factors and litter distribution.
Conclusions:
- Machine learning, specifically ANNs, is a viable and effective approach for assessing marine litter.
- This study provides a novel method for predicting marine litter hotspots and quantities.
- The findings support the use of computational models in marine conservation and management efforts.
