[Scikit-learn-general] New "Sparse- and low-rank approximation wiki",

Hi: There is a new "Sparse- and low-rank approximation wiki", created by Stephen Becker: http://ugcs.caltech.edu/~srbecker/wiki/Main_Page I added a link to the scikit-learn OMP solver: http://ugcs.caltech.edu/~srbecker/wiki/Category:Greedy_Solvers And created a scikit-learn page: http://ugcs.

[Scikit-learn-general] Some searches in the web site point to old versions of scikit-learn

Hi: The outcome of some searches in the website points to old versions of scikit-learn. For instance, if I search for "expectation maximization" I get http://scikit-learn.org/0.5/modules/gmm.html http://scikit-learn.org/0.5/modules/generated/scikits.learn.gmm.GMM.html Note that it points to vers

Re: [Scikit-learn-general] Best classification for very sparse and skewed feature matrix

On Sun, Jan 15, 2012 at 07:39:00PM +0100, Philipp Singer wrote: > The problem is that my representation is very sparse so I have a huge > amount of zeros. That's actually good: some of our estimators are able to use a sparse representation to speed up computation. > Furthermore the dataset is ske

[Scikit-learn-general] Best classification for very sparse and skewed feature matrix

Hey guys! I am currently trying to use the best possible classifier for my task. In my case I have regularly slightly more features than training examples and overall about 5000 features. The problem is that my representation is very sparse so I have a huge amount of zeros. The labels range fr

[Scikit-learn-general] Question about probability estimates in svms

sklearn.svm.SVC has a parameter probability, which enables 'probability' estimates. How are these estimates calculated (% Right during training? Similarity to training data?) ? Thanks, Steven -- RSA(R) Conference 2012 Mar