What is GALSVM?

GALSVM is a free library to analyze morphologies of galaxies based on a particular family of learning machines called support vector machines. The method can be seen as a generalization of the classical CAS classification but with an unlimited number of dimensions and non-linear boundaries between decision regions. It is fully automated and consequently well adapted to large cosmological surveys.

Basically the machine is trained with a simulated sample built from a local visually classified sample chosen in the high-redshift's sample rest-frame and artificially redshifted to match the observing conditions. It has been proven to work properly even on seeing limited data.

The library is written in IDL and uses the free library LibSVM.

The method is fully described in a paper submitted to A&A.


A beta release is now available for download.
CAUTION!!! This is just a preliminary version so it is very possible that you encounter some bugs.
Please let me know (marc.huertas at obspm.fr)!!