[
https://issues.apache.org/jira/browse/MATH-1524?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17058422#comment-17058422
]
Chen Tao edited comment on MATH-1524 at 3/13/20, 5:02 AM:
----------------------------------------------------------
An usecase of "closest":
{code:java}
// I am not sure where could put the closeIndexes properly,
<T extends Clusterable> int[] closestIndexes(List<T> list, DistanceMeasure
dist);
// ...
// ++++++The training phase+++++++
// Use k-means to generate a clusters.
List<? extends CentroidCluster<? extend Clusterable>> clusters = ...
// Extract centers, it is a little strange to let user to do it.
List<? extends Clusterable> centers =
clusters.stream().map(cluster->cluster.centroid()).collect(Collectors.toList());
// -----------End of the training phase-----
// +++++The predict phase++++++++
Clusterable pointToPredict = ...
int[] bestClusterIndexes = aUtilClass.closestIndexes(clusters, measure);
Cluster bestCluster = clusters.get(bestClusterIndexes [0]);
int[] closestPointsIndexes =
aUtilClass.closestIndexes(clusters.getPoints());
Clusterable closestPoints = Arrays.stream(closestPointsIndexes ).map(i ->
bestCluster .getPoints().get(i));
// Other logic like return more information of closestPoints
// ...
// ------------End of the predict phase------------
{code}
was (Author: chentao106):
An usecase of "closest":
{code:java}
// I am not sure where could put the closeIndexes properly,
<T extends Clusterable> int[] closestIndexes(List<T> list, DistanceMeasure
dist);
// ...
List<? extends CentroidCluster<? extend Clusterable>> clusters = ...
List<? extends Clusterable> centers =
clusters.stream().map(cluster->cluster.centroid()).collect(Collectors.toList());
Clusterable pointToPredict = ...
int[] bestClusterIndexes = aUtilClass.closestIndexes(clusters, measure);
Cluster bestCluster = clusters.get(bestClusterIndexes [0]);
int[] closestPointsIndexes =
aUtilClass.closestIndexes(clusters.getPoints());
Clusterable closestPoints = Arrays.stream(closestPointsIndexes ).map(i ->
bestCluster .getPoints().get(i));
// Other logic like return more information of closestPoints
// ...
{code}
> "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
> Attachments: centroid.png, getCenter.png
>
> Time Spent: 20m
> 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)
