Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
[MRG+2] Adding Implementation of SAG - next episode #4738
Actually, newton-cg is not faster.
I tried using the same regularization in SAG, and it converges to the same minimum as liblinear.
Currently, in order to match
referenced this pull request
May 22, 2015
Yes, each data point is chosen randomly, and the same data point can be chosen several times in a row ("draw with replacement").
In Mark Schmidt's presentation (co-author of SAG), slide 76:
I think we should keep it, since 'sag' solver is less efficient if the dataset is small.
Actually, there is two kinds of warm starting for SAG:
If we want warm starting in
some more feedback (I start using it and some some usability issues come to mind).
The function signature of
Thanks a lot for all your reviews !
Changing the signature of
About default alpha value, I changed it to 1, to be consistent with Ridge's default alpha, and LogisticRegression's default C.
About the scaling of alpha by
For example in this benchmark, I need to scale
To avoid confusion, I now changed the named in
As suggested, I also added a boolean parameter