I am working on multiclass problem (num_class=100). I have created a classification model in graphlab/Turi
import graphlab as gl
model= gl.boosted_trees_classifier.create(tr_data, target='species',max_depth=4, max_iterations=500)
Then predicted the probabilities of test set
model.classify(te_data)
Its producing majority class probabilities. How can I modify the model to get probabilities of all 100 classes of target. Thanks in anticipation.
