Regarding scikit-learn, the only thing that we cache is the transformer
processing in the pipeline (see the memory parameter in Pipeline).
It seems that you are passing a different set of features at each
iteration. Is the number of features different?
On Sun, 29 Mar 2020 at 19:23, Pedro Cardoso
I don't think that we rescale the sample_weight and therefore the results
should be different.
On Fri, 24 Apr 2020 at 12:41, Francesco basciani wrote:
> Hi, i have a question regarding the class weights in SVC. I have an
> imbalanced binary classification problem. In my case the ratio between th
The algorithm in scikit-learn-extra are usually algorithms which did not
meet the inclusion criteria (too early publication, not enough citations,
etc.)
However, the code quality is as good and tested than scikit-learn (usually
they were PR in the main repository).
Doing in this manner allows us to
