Le 18/06/12 19:30, Ben Webb a écrit : > On 06/18/2012 05:31 AM, Yannick Spill wrote: >> In the light of my recent discovery of the statistics module in imp, I >> was wondering if some of you would be interested in an implementation of >> self-organizing maps (SOM). They are a generalization of principal >> components analysis to nonlinear models. >> http://en.wikipedia.org/wiki/Self-organizing_map >> There's a shortcut for us to have it in IMP. It's called GPLVM, and is a >> probabilistic SOM based on gaussian processes. Is that something that >> would interest anyone? > > If you're talking about GP-LVM from > http://www.cs.man.ac.uk/~neill/gplvmcpp/ then we can't include that in > IMP, unfortunately, because it is not open source software. > > Ben Sure, but the paper describes how it is made. I already wrote an implementation of the gaussian process, and one would just need to adapt it to this case by writing the covariance function and its derivatives. I'm willing to do it, although I'd really appreciate some programming help to speedup things, but I wasn't going to do it without anyone interested. As far as where to put it, the core of the gaussian process is in the isd module, so no need to rename stuff for the moment.
Y