phanein / deepwalk

DeepWalk - Deep Learning for Graphs
http://www.perozzi.net/projects/deepwalk/
Other
2.68k stars 826 forks source link

A question about how to use the embedding the deepwalk generate #40

Closed songzenghui closed 7 years ago

songzenghui commented 7 years ago

I would like to know how the result of the experiment are produced?Do you use the Liblinear that is mentioned in your paper or use the scoring.py in the deepwalk file? I 'm a newcomer in network representation,and look forward to your reply!

GTmac commented 7 years ago

We use scoring.py to produce the result in the DeepWalk paper. In scoring.py: https://github.com/phanein/deepwalk/blob/master/example_graphs/scoring.py#L94 You can see that the underlying model is sklearn.linear_model.LogisticRegression, which by default uses the LibLinear library (see http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html).