> So if I want to reach like "continue training", I should choose model with
> partial_fit, right?
Yes.
> but I saw nothing have partial_fit function in ensemble methods,
Hm, technically, if the models in the ensemble support partial_fit the ensemble
method itself should also be able to use partial_fit. My guess is that it is
not implemented because it cannot be guaranteed that the individual models
support partial_fit. However, if you are using the voting classifier, you could
probably just train the individual models of the ensemble, because the voting
classifier's decision rule is fixed.
I think the following could work if the estimators_ support partial_fit:
voter = VotingClassifier(...)
voter.fit(...)
For further training:
for i in len(estimators_):
voter.estimators_[i].partial_fit(...)
Best,
Sebastian
> On Feb 1, 2019, at 12:52 AM, lampahome <pahome.c...@mirlab.org> wrote:
>
>
>
> Sebastian Raschka <m...@sebastianraschka.com> 於 2019年2月1日 週五 下午1:48寫道：
> Hi there,
>
> if you call the "fit" method, the learning will essentially start from
> scratch. So no, it doesn't consider previous training results.
> However, certain algorithms are implemented with an additional partial_fit
> method that would consider previous training rounds.
>
> So if I want to reach like "continue training", I should choose model with
> partial_fit, right?
>
> What I want is regression, but I saw nothing have partial_fit function in
> ensemble methods,
>
> Can found in other places?
>
> thx
> _______________________________________________
> scikit-learn mailing list
> scikit-learn@python.org
> https://mail.python.org/mailman/listinfo/scikit-learn
_______________________________________________
scikit-learn mailing list
scikit-learn@python.org
https://mail.python.org/mailman/listinfo/scikit-learn
