Re: [Scikit-learn-general] grid search & x-validation

2011-11-27 Thread al so
Is there an algorithm I can look at to implement my own Grid search and cross validation? On Sun, Nov 27, 2011 at 4:42 AM, Olivier Grisel wrote: > 2011/11/27 al so : > > Is there a Java sample code which performs grid search and > cross-validation > > to find optimum C and Gamma for RBF/C-SVC? >

[Scikit-learn-general] Feature selection criterion

2011-11-27 Thread al so
Say you are working with text document classification in a particular domain. You want to train the system. Is there an established criteria to choose the right Feature vector? -- All the data continuously generated in your

Re: [Scikit-learn-general] Gradient descent backprop on layered RBMs

2011-11-27 Thread David Warde-Farley
A few things I'd add to Olivier's reply: First, it's not quite accurate to call it "layered RBMs". The RBM interpretation, and the CD-1 approximate training procedure, really only make sense in the context of a single layer/unsupervised training, but we then take the weights and biases and shove t

Re: [Scikit-learn-general] A new jenkins integration server is online for scikit-learn

2011-11-27 Thread Olivier Grisel
2011/11/28 Olivier Grisel : > I could not resist: > > I have configured the pep8 report: > >   > https://jenkins.shiningpanda.com/scikit-learn/job/scikit-learn-py27-numpy1.5.1-scipy-0.10.0/ws/pep8.txt The pep8 report is now archived here: https://jenkins.shiningpanda.com/scikit-learn/job/scik

Re: [Scikit-learn-general] Gradient descent backprop on layered RBMs

2011-11-27 Thread Olivier Grisel
You should definitely have a look at theano that will probably run much faster than pure numpy for this kind of models (esp. if you have access to a GPU with the CUDA runtime). http://deeplearning.net/software/theano/ The deep learning tutorial [1] have a section on backpropagation [2] and also o

[Scikit-learn-general] Gradient descent backprop on layered RBMs

2011-11-27 Thread Timmy Wilson
Hi scikit-learn community, I'm experimenting w/ unsupervised Deep Belief Nets (DBN) for dimension reduction. Hinton shows good results using a 2000-500-250-125-2 Autoencoder to cluster a newswire corpus (essentially a neural topic model): http://www.cs.toronto.edu/%7Ehinton/science.pdf I'm tryi

Re: [Scikit-learn-general] A new jenkins integration server is online for scikit-learn

2011-11-27 Thread Olivier Grisel
I could not resist: I have configured the pep8 report: https://jenkins.shiningpanda.com/scikit-learn/job/scikit-learn-py27-numpy1.5.1-scipy-0.10.0/ws/pep8.txt There is currently 511 errors. Please feel free to fix some of them to make the plot decrease here: https://jenkins.shiningpanda.

[Scikit-learn-general] A new jenkins integration server is online for scikit-learn

2011-11-27 Thread Olivier Grisel
Hi, I have just setup a new jenkins instance in the http://shiningpanda.com Continuous Integration cloud service: https://jenkins.shiningpanda.com/scikit-learn/ Right now I have configured a virtualenv with numpy 1.5.1 and scipy 0.10.0 prebuilt on the VM image. It just run the `make` target of

Re: [Scikit-learn-general] scikit test failure on osx

2011-11-27 Thread Robert Kern
On Sun, Nov 27, 2011 at 20:08, Vlad Niculae wrote: > On Sun, Nov 27, 2011 at 10:03 PM, Kenneth C. Arnold > wrote: >> a Mac side note: >> >> I have found that MacPorts solves most of my >> getting-things-running-on-Mac problems. Either you can just use their >> packages directly, often with precom

Re: [Scikit-learn-general] grid search & x-validation

2011-11-27 Thread al so
This is awesome information. Thanks. On Sun, Nov 27, 2011 at 4:42 AM, Olivier Grisel wrote: > 2011/11/27 al so : > > Is there a Java sample code which performs grid search and > cross-validation > > to find optimum C and Gamma for RBF/C-SVC? > > > > Also, any pointers to better java based svm lib

Re: [Scikit-learn-general] scikit test failure on osx

2011-11-27 Thread Vlad Niculae
On Sun, Nov 27, 2011 at 10:03 PM, Kenneth C. Arnold wrote: > a Mac side note: > > I have found that MacPorts solves most of my > getting-things-running-on-Mac problems. Either you can just use their > packages directly, often with precompiled binary downloads, or at > least `port info py27-scipy`

Re: [Scikit-learn-general] scikit test failure on osx

2011-11-27 Thread Kenneth C. Arnold
a Mac side note: I have found that MacPorts solves most of my getting-things-running-on-Mac problems. Either you can just use their packages directly, often with precompiled binary downloads, or at least `port info py27-scipy` will show you the package names for the dependencies. (I'm currently ru

Re: [Scikit-learn-general] grid search & x-validation

2011-11-27 Thread Olivier Grisel
2011/11/27 al so : > Is there a Java sample code which performs grid search and cross-validation > to find optimum C and Gamma for RBF/C-SVC? > > Also, any pointers to better java based svm libraries (besides > libsvm/jlibsvm)? As this is a python project you are likely to get better answers on by

Re: [Scikit-learn-general] scikit test failure on osx

2011-11-27 Thread Vlad Niculae
Hello Massimo I believe this is an issue others, including me, have faced: https://github.com/scikit-learn/scikit-learn/issues/445 https://github.com/scikit-learn/scikit-learn/issues/330 I reverted to the stable versions of numpy and scipy from their website, and the bleeding-edge scikit-learn,

[Scikit-learn-general] grid search & x-validation

2011-11-27 Thread al so
Is there a Java sample code which performs grid search and cross-validation to find optimum C and Gamma for RBF/C-SVC? Also, any pointers to better java based svm libraries (besides libsvm/jlibsvm)? -- All the data continu