[jira] [Created] (SPARK-33518) Improve performance of ML ALS recommendForAll by GEMV

[zhengruifeng (Jira)]

zhengruifeng created SPARK-33518:
------------------------------------

             Summary: Improve performance of ML ALS recommendForAll by GEMV
                 Key: SPARK-33518
                 URL: https://issues.apache.org/jira/browse/SPARK-33518
             Project: Spark
          Issue Type: Improvement
          Components: ML
    Affects Versions: 3.2.0
            Reporter: zhengruifeng


There were a lot of works on improving ALS's {{recommendForAll}}

For now, I found that it maybe futhermore optimized by

1, using GEMV;

2, directly aggregate on topK collections (srcId, Array(dstId), Array(score)), 
instead of each element (srcId, (dstId, score));

3, use guava.ordering instead of BoundedPriorityQueue;

 



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org