[mlpack-git] [mlpack/mlpack] Modeling LSH For Performance Tuning (#749)

[Ryan Curtin]

> + * @article{Dong2008LSHModel,
> + *  author = {Dong, Wei and Wang, Zhe and Josephson, William and Charikar,
> + *      Moses and Li, Kai},
> + *  title = {{Modeling LSH for performance tuning}},
> + *  journal = {Proceeding of the 17th ACM conference on Information and
> + *      knowledge mining - CIKM '08},
> + *  pages = {669},
> + *  url = {http://portal.acm.org/citation.cfm?doid=1458082.1458172},
> + *  year = {2008}
> + * }
> + * @endcode
> + *
> + * We use a different method to fit Gamma Distributions to pairwise distances.
> + * Instead of the MLE method proposed in the paper above, we use the mlpack
> + * class GammaDistribution, which implements fitting according to Thomas Minka's
> + * work.

I think that Minka's method actually does return the MLE (as your convergence criterion tends to 0), so we are still estimating the MLE, just using a clever algorithm.  Correct me if I'm wrong about that.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/pull/749/files/cdcb575826bfb3bd0ef4cafacf465435b3d6d144#r74688501
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160813/6c732ce8/attachment.html>