On Wed, Oct 05, 2011 at 10:29:43AM +0200, Lars Buitinck wrote: > > In the third paragraph of his paper, he mentions:
> > Even when software packages do not use the ineļ¬cient primitive [stepwise, > > procedural] algorithm (as SciPy (Eads,2007) and the R (R Development Core > > Team, 2011) methods hclust and agnes do), the > > author found that implementations largely use suboptimal algorithms rather > > than improved > It would be interesting to experiment Yeah. I wonder how much bullshit there is their. Or how much this is data/application dependent. The reason I say this is that after implementing or fancy heap queue based clustering, we profiled it against the maybe more naive implementation in scipy, and found that the alater was doing much better. G ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity and more. Splunk takes this data and makes sense of it. Business sense. IT sense. Common sense. http://p.sf.net/sfu/splunk-d2dcopy1 _______________________________________________ Scikit-learn-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
