Hi folks, This is my first time contributing to scikit, so please give me a chance if some of my suggestions are not plausible :)
I want to finalize the Multi-Layer Perceptron (MLP) implementation found here: https://github.com/scikit-learn/scikit-learn/pull/1653 I'm thinking of adding two major features: 1) On-line update support that uses the partial_fit() method similar to the one found in Stochastic gradiant descent. It's very helpful for real-time learning and Backpropagation should support online update 2) Weight initialization hyperparameter. This allows the user to set initial weights, since different starter weights for backpropagation could land it to different local optimas, plus, it will give the opportunity to use the weights generated by Restricted Boltzmann Machines as initial weights. So is this a good plan to execute? or do you have other suggestions? Thanks a lot! PS: It would be really helpful if I could know the standard way for pushing code to scikit ------------------------------------------------------------------------------ How ServiceNow helps IT people transform IT departments: 1. A cloud service to automate IT design, transition and operations 2. Dashboards that offer high-level views of enterprise services 3. A single system of record for all IT processes http://p.sf.net/sfu/servicenow-d2d-j _______________________________________________ Scikit-learn-general mailing list Scikit-learn-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
