> So how does mean shift come into the picture? Mean shift exploits this KDE idea by imagining what the points would do if they all climbed up hill to the nearest peak on the KDE surface. It does so by iteratively shifting each point uphill until it reaches a peak.
I wonder if this would still work with high-dimensional data. Distance and space start to act weird in higher dimensions, right?
> The common theme of these problems is that when the dimensionality increases, the volume of the space increases so fast that the available data become sparse. This sparsity is problematic for any method that requires statistical significance.
Although I wonder if you can't account for that by simply increasing the kernel bandwidth. Perhaps not. Perhaps this is why mean-shift seems to be mostly used for computer vision and image processing, where (I guess?) the number of dimensions is low.