May 13, 2025 – We are very proud to announce that our team achieved the highest score in the first machine learning GeoHab 2025 MLWG Competition.
This global competition aimed to promote machine learning in benthic habitat mapping, build consensus on best practices, and provide benchmark datasets for future research. The task was to develop a regression model to predict the mean grain size of seafloor sediment samples on the Scotian Shelf, off Nova Scotia, Canada. Participants used only the official dataset, which included bathymetric, oceanographic, and geospatial variables.
Oikon’s model delivered the most accurate predictions, ranking first among eleven international teams. This result highlights our team’s strong skills in applying machine learning to real-world environmental data, especially in support of offshore wind energy projects.
Congratulations to Bruno, Tea and Luka on this achievement!
More on the competition: GeoHab 2025 MLWG Competition
