Hi Ryan I was going through the "Tree independent Dual tree Algorithm" paper(wrote by you) which talks about the some of the algorithms.
Can you explain how does the tight upper bound of k-nearest neighbor work intuitively? "The basic problem is: given a set of points and integer k, choose k points that minimize the maximum distance between a point in the set and its nearest neighbor in the set" How are you thinking to extend it to Dual tree algorithms? Ankit Aggarwal On Mon, Apr 24, 2017 at 9:02 PM, Ryan Curtin <[email protected]> wrote: > On Sun, Apr 23, 2017 at 07:21:34PM +0530, Ankit Aggarwal wrote: > > Hi Ryan > > > > I was able to found the following thread talking about "Fast K-Centre > > Algorithm" > > > > "Yes, this project would be to design an algorithm, and then spend the > > summer implementing it. Unfortunately, Google funds students for > > implementation and not for research," > > I got to know through the links that dual-tree algorithm is implemented > in > > src/mlpack/methods. So I would be going through out the src package > defines > > above. > > > > Can you please share what should be the timeline for this project? > > Hi Ankit, > > I am not sure what the timeline should be since I don't know what > obligations you have or how fast you might want to work on this. So > that part is up to you. > > Please be aware though that I don't expect to have much time this summer > to try to help, so if you are going to work on the k-centers idea, you > will have to be pretty self-sufficient, although I will help where I > can. I would strongly suggest reading all of the papers linked to in > the other mailing list threads, since like you pointed out this is a > research project involving the design of an original algorithm. > > Thanks, > > Ryan > > -- > Ryan Curtin | "So I just, uh... I just cut them up like regular > [email protected] | chickens?" - Henry Spencer >
_______________________________________________ mlpack mailing list [email protected] http://knife.lugatgt.org/cgi-bin/mailman/listinfo/mlpack
