You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by "Zhipeng Zhang (Jira)" <ji...@apache.org> on 2022/11/01 04:18:00 UTC

[jira] [Created] (FLINK-29824) AgglomerativeClustering fails when the distanceThreshold is very large

Zhipeng Zhang created FLINK-29824:
-------------------------------------

             Summary: AgglomerativeClustering fails when the distanceThreshold is very large
                 Key: FLINK-29824
                 URL: https://issues.apache.org/jira/browse/FLINK-29824
             Project: Flink
          Issue Type: Bug
          Components: Library / Machine Learning
    Affects Versions: ml-2.2.0
            Reporter: Zhipeng Zhang


The current implementation did not consider following case:

When distanceThreshold not null and set as a large value, all data points are supposed to be assigned into a single cluster.

 

In this case, we should stop training when the number of the active clusters is one.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)