Hi,

Have you tried to ask the authors for their source code?

If your implementation is based on scikit-learn, it could be interesting to
share it as a gist.

I had a quick look at the paper and it seems to me that you don't even need
to change liblinear. You just need to prepare a sparse matrix X of shape
[n_samples * n_clusters, n_features * (n_clusters + 1)] and a y vector of
shape [n_samples * n_clusters]. Then you can fit a LinearSVC with X and y.
Of course, you do need to re-implement the predict method but no direct
changes to liblinear are needed.

M.


On Mon, May 26, 2014 at 10:14 AM, Kathy Hida <[email protected]> wrote:

> Hi,
>
> Is there any implementation for clustered SVM proposed in this paper
> http://jmlr.org/proceedings/papers/v31/gu13b.pdf.
>
> The implementation seems to be trivial, as we just need to do simple
> modifications on LibLinear. I did this modifications but I still get
> results different from those shown in the paper.
>
>
>
>
>
> ------------------------------------------------------------------------------
> The best possible search technologies are now affordable for all companies.
> Download your FREE open source Enterprise Search Engine today!
> Our experts will assist you in its installation for $59/mo, no commitment.
> Test it for FREE on our Cloud platform anytime!
>
> http://pubads.g.doubleclick.net/gampad/clk?id=145328191&iu=/4140/ostg.clktrk
> _______________________________________________
> Scikit-learn-general mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
>
>
------------------------------------------------------------------------------
The best possible search technologies are now affordable for all companies.
Download your FREE open source Enterprise Search Engine today!
Our experts will assist you in its installation for $59/mo, no commitment.
Test it for FREE on our Cloud platform anytime!
http://pubads.g.doubleclick.net/gampad/clk?id=145328191&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to