srowen commented on issue #27758: [SPARK-31007][ML] KMeans optimization based on triangle-inequality URL: https://github.com/apache/spark/pull/27758#issuecomment-594055650 If I'm reading that right, it's slower to train in many cases in the first instance, and slightly faster at scale in the second. I'm surprised because I'd have thought it's more easily a win, even with the overhead of calculating stats. Is the purpose more about prediction speed? I just wonder how much one is worth vs the other.
---------------------------------------------------------------- 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]
