1) Is there some way to generate C code which would give me floating point probabilities for every prediction category instead of a binary result? I am using the ExtraTreeModel from sklearn (code snippet shown below).
2) My model is almost 1Gb in size, the C code generated is about the same size, which will take forever to compile, is there a way to compress this model or split up the files?
Training Code Snippet
from sklearn.ensemble import ExtraTreesClassifier
from joblib import dump, load
clf = ExtraTreesClassifier(n_estimators=250, random_state=0,verbose=True)
clf.fit(train_x, train_y)
prob_test_data = clf.predict_proba(test_x)
1) Is there some way to generate C code which would give me floating point probabilities for every prediction category instead of a binary result? I am using the ExtraTreeModel from sklearn (code snippet shown below).
2) My model is almost 1Gb in size, the C code generated is about the same size, which will take forever to compile, is there a way to compress this model or split up the files?
Training Code Snippet
Thanks in advance