Closed sevenguin closed 6 years ago
Hi @sevenguin , Thanks for you comment. I went back to the documentation and indeed setting max_features to 1 makes the classifier a totally random decision tree see documentation. I am now confused about what the deep forest authors meant by "completely random forest". I'll go back to the paper and make corrections if I misunderstood this part.
Thanks again for the feedback
@sevenguin , I have double checked the paper and the scikit documentation and you're right about the max_features. It's now corrected and updated. Thanks for noticing!!!
There are two types of forest in the paper, and the first is completely-random tree forests. I think the forest is show in code
crf = RandomForestClassifier(n_estimators=n_tree, max_features=None, min_samples_split=min_samples, oob_score=True, n_jobs=n_jobs)
, in functionwindow_slicing_pred_prob
, if i'm right, then the parameter max_features should be 1.