You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Come up with a clustering (unsupervised learning) machine learning algorithm to cluster vessels by behaviour instead of having set categories, and check how that overlaps with our current categories
Treat the output of the classifier as a likelihood of a vessel being of a certain gear type, where it can simultaneously be of multiple gear types, the likelihoods adding up to 1. Then run the fishing/non-fishing classifiers for all the gear types, and use a weighted average of their scores (multiplying their "is fishing" score by the likelihood that the vessel is of that gear type).
The text was updated successfully, but these errors were encountered:
The text was updated successfully, but these errors were encountered: