What is it about?
The article describes how we used machine learning algorithms inspired by natural selection to obtain predictive equations for the dimensions of underwater dunes. The obtained predictors outperform the relationships typically used by practitioners and based on simple parameters.
Featured Image
Photo by Juan chavez on Unsplash
Why is it important?
It is very important to be able to predict the sizes of dunes at equilibrium for many applications such as navigation, interactions with offshore wind farms, the reconstruction of paleoclimates, the management of sand resources, or the calculation of roughness in circulation models. This study represents a breakthrough compared to previous studies in the performance of predictors.
Perspectives
This research work has been an opportunity for me to discover the performance of genetic algorithms, and I have been fascinated by the way they evolve generations of equations by following the principles of natural selection to retain only the best 'individuals'.
Dr Arnaud Doré
Universite de Bordeaux
Read the Original
This page is a summary of: A hybrid, genetic programming and physically-based predictor of dune geometry, Geomorphology, January 2025, Elsevier,
DOI: 10.1016/j.geomorph.2024.109495.
You can read the full text:
Contributors
The following have contributed to this page
