What is it about?
Advanced clustering methods are required to detect agglomerations of data points of specific shapes. We show how such method can be implemented and used for the detection of eddies in ocean currents. Our method allows oceanographers to quantify the transport of energy and salinity to the Atlantic Ocean more precisely.
Featured Image
Photo by Andrzej Kryszpiniuk on Unsplash
Why is it important?
We show how yet unsolved challenges regarding false positives can be addressed in our use case. Shape-sensitivity allows to go one step further compared to the state-of-the-art density-based clustering, excluding areas of high density lacking specific requirements regarding velocity, thus shape.
Perspectives
The introduced method shows how the sum of multiple small, simple techniques allow to achieve a greater solution to a challenge. It was a pleasure to walk this path throughout all these steps with all co-authors.
Nelson Tavares de Sousa
Christian-Albrechts-Universitat zu Kiel
Read the Original
This page is a summary of: VoCC: Vortex Correlation Clustering Based on Masked Hough Transformation in Spatial Databases, August 2023, ACM (Association for Computing Machinery),
DOI: 10.1145/3609956.3609971.
You can read the full text:
Resources
Contributors
The following have contributed to this page
