[mlpack] contribute to mean-shift clustering

Ryan Curtin gth671b at mail.gatech.edu
Wed Jan 7 09:53:20 EST 2015


On Wed, Jan 07, 2015 at 02:57:55PM +0800, Shangtong Zhang wrote:
> Hi,
> 
> 
> I compiled mlpack from source, run some demo apps and build a development environment in xcode.
> To start, I trace kmeans step by step.
> Now, I want to contribute to the issue Mean-shift clustering method #352.
> I wonder if this feature will be accepted if I just implement basic MS algorithm rather than DTMS.
> 
> 
> I am looking forward to any feedbacks.

Hi Shangtong,

I'm happy to accept it as long as the code adheres to the style
guidelines [1], and outperforms scikit-learn's implementation [2] and
this MATLAB implementation [3], or is at least competitive with it.
Once your implementation is done, I can take a look over it and see if
there are any optimizations we can make.

Thanks!

Ryan

[1] https://github.com/mlpack/mlpack/wiki/StyleGuidelines
[2] http://scikit-learn.org/stable/modules/generated/sklearn.cluster.MeanShift.html
[3] http://coewww.rutgers.edu/riul/research/code.html

-- 
Ryan Curtin    | "Then they attacked a town.  A small town, but a town
ryan at ratml.org | nonetheless.  A town of people.  People who died."


More information about the mlpack mailing list