Le 23/01/2012 11:34, Olivier Grisel a écrit : > 2012/1/23 Andreas<[email protected]>: >> On 01/23/2012 11:28 AM, Adrien wrote: >>> Hello everyone, >>> >>> A quick question: why not use Nystrom instead? >>> >> That was on my GSoC wish list ;) >> The application I had in mind at the moment was >> the label propagation and maybe the spectral clustering. >> In general, I think the thresholding would work >> whenever rbf is used as similarity measure. >> >> For kernel methods, we definitely want the Nystrom method :) > I had never heard of it before. I learn new stuff every day on this > list. Any other reference your recommend or is the 2004 paper by > Fowlkes at al still the state of the art? I would say Fowlkes et al. is the reference paper for Nystrom in the context of spectral clustering. However, the part on indefinite kernels is a bit puzzling to me...
Adrien ------------------------------------------------------------------------------ Try before you buy = See our experts in action! The most comprehensive online learning library for Microsoft developers is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3, Metro Style Apps, more. Free future releases when you subscribe now! http://p.sf.net/sfu/learndevnow-dev2 _______________________________________________ Scikit-learn-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
