A plugin for the GATE language technology framework for training and using machine learning models. Currently supports Mallet (MaxEnt, NaiveBayes, CRF and others), LibSVM, Scikit-Learn, Weka, and DNNs through Pytorch and Keras.
According to the documentation, -b is enabled by default, so these should be probabilities,
but the values shown do not look plausible.
Maybe for LibSVM we do not correctly produce the whole array of probabilities and always take the probability of the same class?
According to the documentation, -b is enabled by default, so these should be probabilities, but the values shown do not look plausible. Maybe for LibSVM we do not correctly produce the whole array of probabilities and always take the probability of the same class?