[
https://issues.apache.org/jira/browse/MATH-1524?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17056204#comment-17056204
]
Gilles Sadowski edited comment on MATH-1524 at 3/10/20, 6:42 PM:
-----------------------------------------------------------------
There was [a discussion on the "dev"
ML|https://markmail.org/message/wotd3mnvperhblnl] where I argued that the
{{CentroidCluster}} class was problematic.
Class {{Cluster}} contains a {{centroid()}} method that will return a value
consistent with the state of the {{Cluster}} instance.
was (Author: erans):
There was [a discussion on the "dev"
ML|https://markmail.org/message/wotd3mnvperhblnl] where I argued that the
{{CentroidCluster}} class was problematic.
Class {{Cluster}} contains a {{centroid()}} method that will be return a value
consistent with the state of the {{Cluster}} instance.
> "chooseInitialCenters" should move out from KMeansPlusPlusClusterer
> -------------------------------------------------------------------
>
> Key: MATH-1524
> URL: https://issues.apache.org/jira/browse/MATH-1524
> Project: Commons Math
> Issue Type: Improvement
> Reporter: Chen Tao
> Priority: Major
> Time Spent: 10m
> Remaining Estimate: 0h
>
> There are two reason for "chooseInitialCenters" should be move out from
> "KMeansPlusPlusClusterer":
> # k-means++ clusterer is a special case of k-means clusterer, that k-means++
> initialize the cluster centers with k-means++ algorithm. Another case is
> initialize the cluster centers with random points.
> # The mini batch k-means will reuse "chooseInitialCenters".
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
