Greg's House of Treats

An excellent resource for compilers.

http://www.math.utah.edu/faq/compilers/compilers.html

Apparently Amarok can copy files to the FUZE.  The trick is the little unlabeled button on the FUZE bar of the Amarok collection list.

http://www.goodbyemicrosoft.net/news.php?item.615.2

After my frustration with LIBSVM, I decided to look into other methods.  Weka suggests that PLS works well for my problem.  But anyway, I need to find a good package (weka is java therefore unsuitable).

http://www.support-vector-machines.org/SVM_soft.html
http://www.linux.org/docs/ldp/howto/AI-Alife-HOWTO-2.html
http://torch5.sourceforge.net/
http://mloss.org/software/
https://spreadsheets.google.com/ccc?key=0Aunb9cCVAP6NdDVBMzY1TjdPcmx4ei1EeUZNNGtKUHc&hl=en#gid=0

Reviews.

Weka
Very good for exploring data.  I found several good algorithms (mostly linear algorithms such as PLS).  However, not a final solution because it is written in Java.

LIBSVM
Performs poorly on my data.

LIBLinear
Not yet benchmarked.

Waffles
This will be good, but currently doesn't compile due to being 64-bit clean (gcc 4.5).

dlib
Looks promising.  There is no benchmarking tool (?)

Shark
Looks promising.  Not yet benchmarked.

Shogun
I previously reviewed this, using the Octave plugin.  However, I don't remember the results.  Need to re-benchmark.