[scikit-learn] RidgeCV with multiple targets returns a single alpha. Is it possible to get one alpha per target?

Hello, I'd like to use RidgeCV to find the optimal alpha for each colunm (ntargets) of the DV variable. It lloks like itthe fit() computes a single alpha. Is there a way to compute one alpha per column? -- -- Christophe Pallier INSERM-CEA Cognitive Neuroimaging Lab, Neurospin, bat 145, 9119

Re: [scikit-learn] RidgeCV with multiple targets returns a single alpha. Is it possible to get one alpha per target?

> On Aug 7, 2018, at 6:43 AM, Christophe Pallier wrote: > > Hello, > > I'd like to use RidgeCV to find the optimal alpha for each colunm (ntargets) > of the DV variable. > > It lloks like itthe fit() computes a single alpha. Is there a way to compute > one alpha per column? > > > > > --

Re: [scikit-learn] RidgeCV with multiple targets returns a single alpha. Is it possible to get one alpha per target?

you should call RidgeCV on all targets separately. HTH Alex On Tue, Aug 7, 2018 at 12:46 PM Christophe Pallier wrote: > > Hello, > > I'd like to use RidgeCV to find the optimal alpha for each colunm (ntargets) > of the DV variable. > > It lloks like itthe fit() computes a single alpha. Is there

[scikit-learn] Normalize an image with 3 channels

Hi everyone, I'm trying to extract features from images using VGG and I have to normalize each image array before doing that. Each array is of the size (224, 224 ,3) and I'm unable to use the MinMaxScalar in this case. Would appreciate any help with this . Thank you ! -- Regards, Pra

Re: [scikit-learn] RidgeCV with multiple targets returns a single alpha. Is it possible to get one alpha per target?

You can get one alpha per target in the Ridge estimator (without CV). Then you would have to code the cv loop yourself. Depending on how many target you have this can be more efficient than looping over targets as Alex suggests. Either way there is some coding to do unfortunately. Michael On

[scikit-learn] Small suggestion for documentation

Dear maintainers,I've just known scikit-learn and found it very useful. Congratulations for this library. I found some confuse terms to describe r2_score parameters in documentation [1]. For me, the meanings of y_true and y_pred are not clear. From [1]:- y_true: ... Ground truth (correct) target

Re: [scikit-learn] Small suggestion for documentation

I think that the vocabulary mismatch comes from the fact that you are looking at these terms thinking about in sample statistics, while they are used here in the context of prediction. I think that in the context of prediction, these are the right terms. Cheers, Gaël ⁣Sent from my phone. Plea

Re: [scikit-learn] Small suggestion for documentation

Hi Gaël, > Em terça-feira, 7 de agosto de 2018 15:24:00 BRT, Gael Varoquaux escreveu: > > > I think that the vocabulary mismatch comes from the fact that you are > looking at these terms thinking about in sample statistics, while they are > used here in the context of prediction. I thi