Hi Christian.
Some clustering algorithm (for example KMeans and MiniBatchKMeans)
provide a "predict" function, which you can use to label new data.
Not all algorithms support this as there is not always an obvious way
to assign labels to new points (for example I don't know of any using
mean-shift).
Best,
Andy

Am 09.11.2012 15:46, schrieb Christian:
> Hi,
>
> after fitting a clusterer  I'll label new data.  Is there an easier way
> instead of building an ex-post classifier.
>
> Many thanks
> Christian
>
>
> example in weka:
> #Building the clusterer and save the object in cluster.cla
> java -cp weka.jar weka.clusterers.EM -t data0.arff -d cluster.cla
>
> === Clustering stats for training data ===
>
> Clustered Instances
> 0       90 ( 12%)
> 1      174 ( 23%)
> 2      204 ( 27%)
> 3        1 (  0%)
> 4      143 ( 19%)
> 5       35 (  5%)
> 6      121 ( 16%)
>
> #Based on custer.cla I could label new data with:
> java -cp weka.jar weka.clusterers.EM -T data1.arff -l cluster.cla -p 0
> 0 1
> 1 4
> 2 1
> 3 4
> 4 0
> 5 4
> 6 4
> 7 5
> 8 0
> 9 6
>
>
>
>
>
>
>
>
>
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_nov
> _______________________________________________
> Scikit-learn-general mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general


------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_nov
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to