Re: [scikit-learn] MLPClassifier as a feature selector

I agree with Gael on this one and am happy to help with the PR if you need any assistance. Best, Maciek Pozdrawiam, | Best regards, Maciek Wójcikowski mac...@wojcikowski.pl 2017-12-29 18:14 GMT+01:00 Gael Varoquaux : > I think that a transform method

Re: [scikit-learn] MLPClassifier as a feature selector

I think that a transform method would be good. We would have to add a parameter to the constructor to specify which layer is used for the transform. It should default to "-1", in my opinion. Cheers, Gaël ⁣Sent from my phone. Please forgive typos and briefness.​ On Dec 29, 2017, 17:48, at

Re: [scikit-learn] MLPClassifier as a feature selector

Hi Thomas, it is possible to obtain the activation values of any hidden layer, but the procedure is not completely straight forward. If you look at the code of the `_predict` method of MLPs you can see the following: ```python def _predict(self, X): """Predict using the trained model

Re: [scikit-learn] MLPClassifier as a feature selector

Alright, with these attributes I can get the weights and biases, but what about the values on the nodes of the last hidden layer? Do I have to work them out myself or there is a straightforward way to get them? On 7 December 2017 at 04:25, Manoj Kumar wrote: >

Re: [scikit-learn] MLPClassifier as a feature selector

Hi, The weights and intercepts are available in the coefs_ and intercepts_ attribute respectively. See https://github.com/scikit-learn/scikit-learn/blob/a24c8b46/sklearn/neural_network/multilayer_perceptron.py#L835 On Wed, Dec 6, 2017 at 4:56 PM, Brown J.B. via scikit-learn <

Re: [scikit-learn] MLPClassifier as a feature selector

I am also very interested in knowing if there is a sklearn cookbook solution for getting the weights of a one-hidde-layer MLPClassifier. J.B. 2017-12-07 8:49 GMT+09:00 Thomas Evangelidis : > Greetings, > > I want to train a MLPClassifier with one hidden layer and use it as a >

[scikit-learn] MLPClassifier as a feature selector

Greetings, I want to train a MLPClassifier with one hidden layer and use it as a feature selector for an MLPRegressor. Is it possible to get the values of the neurons from the last hidden layer of the MLPClassifier to pass them as input to the MLPRegressor? If it is not possible with