[ 
https://issues.apache.org/jira/browse/MATH-959?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Thomas Neidhart updated MATH-959:
---------------------------------

    Attachment: MATH-748.zip

First draft of a Hierarchical Agglomerative Clusterer, similar to the linkage 
method of octave/matlab.

The simple implementation with different linkage criteria is trivial but has 
O(n^3) runtime complexity.

The first version implements the optimal slink algorithm for single-link in 
O(n^2).

There exists also an optimal algorithm for complete-link.

For other criteria, e.g. average, centroid, there is an algorithm with O(n^2 
log n) runtime based on priority queues.
                
> Add HAC clustering algorithm
> ----------------------------
>
>                 Key: MATH-959
>                 URL: https://issues.apache.org/jira/browse/MATH-959
>             Project: Commons Math
>          Issue Type: Sub-task
>            Reporter: Thomas Neidhart
>             Fix For: 4.0
>
>         Attachments: MATH-748.zip
>
>
> Add at least one hierarchical clustering algorithm.
> With the refactoring of the clustering package, this should now be feasible 
> as the Cluster class can be extended.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to