[scikit-learn] clustering on big dataset

[Shiheng Duan]

Thanks, Joel,
I am working on KD-tree to find the nearest neighbors. Basically, I find
the nearest neighbors for each point and then merge a couple of points if
they are both NN for each other. The problem is that after each iteration,
we will have a new bunch of points, where new clusters are added. So the
tree needs to be updated. Since I didn't find any dynamic way to update the
tree, I just rebuild it after each iteration, costing lots of time. Any
idea about it?
Actually, it takes around 16 mins to build the tree in the first iteration,
which is not slow I think. But it still runs slowly. I have a dataset of
12*872505 (features, samples). It takes several days to run the program. Is
there any way to speed up the query process of NN? I doubt query may be too
slow.
Thanks for your time.

On Thu, Jan 4, 2018 at 3:55 AM, Joel Nothman <joel.nothman at gmail.com> wrote:

> Can you use nearest neighbors with a KD tree to build a distance matrix
> that is sparse, in that distances to all but the nearest neighbors of a
> point are (near-)infinite? Yes, this again has an additional parameter
> (neighborhood size), just as BIRCH has its threshold. I suspect you will
> not be able to improve on having another, approximating, parameter. You do
> not need to set n_clusters to a fixed value for BIRCH. You only need to
> provide another clusterer, which has its own parameters, although you
> should be able to experiment with different "global clusterers".
>
> On 4 January 2018 at 11:04, Shiheng Duan <shiduan at ucdavis.edu> wrote:
>
>> Yes, it is an efficient method, still, we need to specify the number of
>> clusters or the threshold. Is there another way to run hierarchy clustering
>> on the big dataset? The main problem is the distance matrix.
>> Thanks.
>>
>> On Tue, Jan 2, 2018 at 6:02 AM, Olivier Grisel <olivier.grisel at ensta.org>
>> wrote:
>>
>>> Have you had a look at BIRCH?
>>>
>>> http://scikit-learn.org/stable/modules/clustering.html#birch
>>>
>>> --
>>> Olivier
>>> ​
>>>
>>> _______________________________________________
>>> scikit-learn mailing list
>>> scikit-learn at python.org
>>> https://mail.python.org/mailman/listinfo/scikit-learn
>>>
>>>
>>
>> _______________________________________________
>> scikit-learn mailing list
>> scikit-learn at python.org
>> https://mail.python.org/mailman/listinfo/scikit-learn
>>
>>
>
> _______________________________________________
> scikit-learn mailing list
> scikit-learn at python.org
> https://mail.python.org/mailman/listinfo/scikit-learn
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/scikit-learn/attachments/20180104/be52caa9/attachment-0001.html>