zhengruifeng opened a new pull request #26948: [SPARK-30120][ML] LSH approxNearestNeighbors should use BoundedPriorityQueue when numNearestNeighbors is small URL: https://github.com/apache/spark/pull/26948 ### What changes were proposed in this pull request? Use Top-K aggregation instead of approxNearestNeighbors ### Why are the changes needed? performance improvement: 1, avoid count job to estimate `quantile` 2, exact threshold, make filtered dataset `modelSubset` probably smaller ### Does this PR introduce any user-facing change? No ### How was this patch tested? existing testsuite & manually performance test
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
