Github user NathanHowell commented on the pull request:
https://github.com/apache/spark/pull/8246#issuecomment-139345239
@jkbradley it's not that uncommon to use sparse features for time series
classification (sparse interval features) with RF.
training a linear model with spark.ml's pipeline is (was?) prohibitively
expensive for this dataset. normalizing features using the spark.ml's
transforms generates two stages per feature... rather than two stages total
(the first for density estimation and a second for the bucket transform) but
that sort of defeats the purpose of using `ml` instead of `mllib`..
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
