« first day (1927 days earlier)      last day (3073 days later) » 

Glen_b
Glen_b
1:48 AM
@Antoni it's okay I deleted it myself once I read it
 
 
5 hours later…
Dawny33
Dawny33
6:37 AM
Hey folks. Can someone explain this?
1
Q: Why Result Dithering?

Student TMost of us want to build a model as accurate as possible, however, an experienced chief data scientist believes a practical machine learning algorithm should randomly shuffle the results, therefore non-optimal results. This is known as "Results Dithering". Slide 15 at: http://cikm2013.org/slides...

machine-learning
 
 
6 hours later…
Andre Silva
Andre Silva
12:40 PM
Does anyone know a key reference for citing the kmeans clustering technique: en.wikipedia.org/wiki/K-means_clustering? I want to cite this method, but I am not sure what would be a good scientific article to reference it. I'm using it in the context of clustering univariate data (for class interval definition) for map representation. Tks.
 
Nick Cox
Nick Cox
1:38 PM
@AndreSilva What looks to be an excellent historical survey at jehps.net/Decembre2008/Bock.pdf
 
 
1 hour later…
Scortchi
Scortchi
3:06 PM
@Dawny33: Since you drew our attention to the question it's been edited to reveal that the context is recommendation engines. So perhaps it's just that you don't want to keep showing someone the same thing again & again.
 
 
2 hours later…
Andre Silva
Andre Silva
4:37 PM
@NickCox, I'll use it. Thank you very much!
 
 
4 hours later…
gung
gung
9:06 PM
@AndreSilva, here is a classic w/ lots of references: [Data clustering: 50 years beyond K-means](http://www.sciencedirect.com/science/article/pii/S0167865509002323). It says k-means "was
independently discovered in different scientific fields by Steinhaus (1955), Lloyd (1957), Ball & Hall (1965) and McQueen (1967)".
 
 
2 hours later…
Andre Silva
Andre Silva
11:32 PM
It seems awesome. Thanks @gung.
 

« first day (1927 days earlier)      last day (3073 days later) »