4 ms·
The crosscat paper (http://web.mit.edu/vkm/www/crosscat.pdf http://web.mit.edu/vkm/www/crosscat.pdf) is super-terse. Is there a more gentle description somewher
by murbard2 13y ago
The crosscat paper (http://web.mit.edu/vkm/www/crosscat.pdf http://web.mit.edu/vkm/www/crosscat.pdf) is super-terse. Is there a more gentle description somewhere?
I'm quite familiar with generative models and MCMC sampling and reasonably familiar with the Dirichlet process though I've never implemented it.
- jaybaxter 13y agoThis computational cognitive science paper about CrossCat (http://shaftolab.com/lab_papers/repository/shaftoetal06_crosscat.pdf http://shaftolab.com/lab_papers/repository/shaftoetal06_cros...) provides a different perspective that may be more approachable.