Skip to content
Advertisement

Backpropagation with Momentum using Scikit-Learn

I’m trying to use Scikit-Learn’s Neural Network to classify my dataset using a Backpropagation with Momentum. I need to specify these parameters: Hidden neurons, Hidden layers, Training set, Learning rate and Momentum. I found MLPClassifier in Sklearn.neural_network package. The problem is that this package is part of Scikit-learn V0.18 which is a dev version.

Is there a way I could use Scikit-Learn V0.17 to do this?

Using Anaconda, but I can change that if it causes problems.

Advertisement

Answer

If anybody ever needs an answer for this, I actually decided to run everything on a Linux VM. I then followed the instructions to install the dev version and everything(well almost) worked perfectly. Running it on Linux is way easier than Windows because you can just install the package from git and run it without having to download required software to compile it. I still struggled a little bit though.

User contributions licensed under: CC BY-SA
7 People found this is helpful
Advertisement