Orange Forum • View topic - k-means - performance measure

k-means - performance measure

A place to ask questions about methods in Orange and how they are used and other general support.

k-means - performance measure

Postby bonjovi2012 » Sun Jul 15, 2012 16:36

Hello Experts,

When I want to evaluate the performance of a classification task in Orange, I use the overall accuracy of a model.

Similarly, for regression tasks, I use the RMSE to determine the level of performance achieved by a regression model.

I am wondering what performance measures can I use for evaluating the performance of clustering tasks (e.g. kmeans)?

thank you

Re: k-means - performance measure

Postby bonjovi2012 » Sun Jul 15, 2012 16:38

P.S. I forgot to add that I want to do this analysis in Orange Canvas as I don't have the time to learn Phython.

Cheers

Re: k-means - performance measure

Postby bonjovi2012 » Thu Aug 02, 2012 22:43

Hi,

I see that I should be using a measure such as Davies Bouldin. Can Orange calculate the Davies Bouldin for a kmeans model?

Thanks,
Colm

Re: k-means - performance measure

Postby Ales » Fri Aug 03, 2012 10:37

bonjovi2012 wrote:I see that I should be using a measure such as Davies Bouldin. Can Orange calculate the Davies Bouldin for a kmeans model?

No. This measure is not implemented.


Return to Questions & Support