Hi,
You could try using MiniBatchKMeans and feed the generator to the
partial_fit method.
On Thu, Oct 30, 2014 at 4:37 PM, Dzung Nguyen wrote:
> I'm running bag of words model on a big dataset of images. It generates a
> huge raw feature matrix ( before doing kmeans to get codebook) which can't
I'm running bag of words model on a big dataset of images. It generates a
huge raw feature matrix ( before doing kmeans to get codebook) which can't
be stored in memory. Can I use generator/yield in python and feed that to
kmeans object?
Thanks,
--
*Dzung Nguyen*
PhD Student
Electrical Engineeri
