lucasb-eyer / pydensecrf

Python wrapper to Philipp Krähenbühl's dense (fully connected) CRFs with gaussian edge potentials.
MIT License
1.94k stars 413 forks source link

How to get the hyperparameters? #54

Closed lixiang-ucas closed 6 years ago

lixiang-ucas commented 6 years ago

Thanks for you wonderful work. I have trying inferring using your code, but it works not good. I'm wondering how to find the best hyperparameters? Have you implemented the learning procedure?

lucasb-eyer commented 6 years ago

I have not implemented the learning procedure, but am happy to accept PRs implementing it, as mentioned in the README.

Besides that, finding hyperparameters is an open question in ML, so nothing I can help with in a github issue :smile: Use a validation-set and do search.