[mlpack-git] [mlpack] Mean shift clustering (#388)

[Shangtong Zhang]

I make some changes to make my implementation in strict accordance with the paper above.
In the paper, only radially symmetric kernels are referred, 
so the template parameter metric won't make sense.
I think only Euclidean Distance is meaningful according to the deduction.
And in the past I use all points to calculate new centroid, 
now I only use points near the current centroid according to the paper.
I implement Gradient() in Gaussian Kernel as dk(s)/ds.
If it's appropriate, I will implement Gradient() in all radially symmetric kernels next.

---
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/pull/388#issuecomment-72316708
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20150131/d95e790f/attachment.html>