On Wed, Oct 30, 2013 at 12:49:49AM +0100, Vlad Niculae wrote:
> Adding "L1" (elementwise) regularization makes L-BFGS-B converge much
> quicker. This is cool because for NMF such a penalty has other
> advantages.
By the way, the MiniBatchDictLearning can be trivially modified to do
this: do a non
On 10/29/2013 07:19 AM, jim vickroy wrote:
On 10/29/2013 5:11 AM, Olivier Grisel wrote:
2013/10/23 j vickroy:
On 10/23/2013 10:18 AM, Andreas Mueller wrote:
FYI, the features I would use for the superpixel based approach would be
"color" histogramms (bag of words of channel intensities):
Resha
Hi all,
During the PyCon sprint I kept digging into the NMF and specifically
ways to solve each sub-iteration. It became clear that the alternating
NLS approach finds good reconstructions and converges well, but the
NLS solving step is critical and must be optimized.
I have started looking into d
On 10/29/2013 5:11 AM, Olivier Grisel wrote:
2013/10/23 j vickroy :
On 10/23/2013 10:18 AM, Andreas Mueller wrote:
FYI, the features I would use for the superpixel based approach would be
"color" histogramms (bag of words of channel intensities):
Reshape the images to (-1, 6) so you have lists
2013/10/23 j vickroy :
> On 10/23/2013 10:18 AM, Andreas Mueller wrote:
>> FYI, the features I would use for the superpixel based approach would be
>> "color" histogramms (bag of words of channel intensities):
>> Reshape the images to (-1, 6) so you have lists of pixels (subsample if
>> they are ma
On 10/23/2013 10:18 AM, Andreas Mueller wrote:
> FYI, the features I would use for the superpixel based approach would be
> "color" histogramms (bag of words of channel intensities):
> Reshape the images to (-1, 6) so you have lists of pixels (subsample if
> they are many), run (MiniBatch)KMeans, a
2013/10/28 Marcel Caraciolo :
> Thanks Lard and Jaques we are still at development. We are referencing some
> distance metrics of Scikit-learn and unfortunatelly The author missed to
> cross-reference the code for the scikit. Apologize ourserves, and we are
> still working on it. People abandoned t
