KMeans clustering algorithm (apache#13604)

Implement Kmeans clustering algorithm for vectors. Knn algorithms that further reduce memory usage of vectors (such as Product Quantization, RaBitQ etc) require clustering of vectors. This implements KMeans clustering algorithm. Co-authored-by: Jim Ferenczi [email protected]
kaivalnp · Aug 2, 2024 · 26b46ce · 26b46ce
1 parent cb8bc75
commit 26b46ce
Show file tree

Hide file tree

Showing 6 changed files with 692 additions and 0 deletions.
diff --git a/lucene/CHANGES.txt b/lucene/CHANGES.txt
@@ -137,6 +137,8 @@ New Features
 * GITHUB#13597: Align doc value skipper interval boundaries when an interval contains a constant
   value. (Ignacio Vera)
 
+* GITHUB#13604: Add Kmeans clustering on vectors (Mayya Sharipova, Jim Ferenczi, Tom Veasey)
+
 Improvements
 ---------------------
 

diff --git a/lucene/sandbox/src/java/module-info.java b/lucene/sandbox/src/java/module-info.java
@@ -22,6 +22,7 @@
 
   exports org.apache.lucene.payloads;
   exports org.apache.lucene.sandbox.codecs.idversion;
+  exports org.apache.lucene.sandbox.codecs.quantization;
   exports org.apache.lucene.sandbox.document;
   exports org.apache.lucene.sandbox.queries;
   exports org.apache.lucene.sandbox.search;