Github user jkbradley commented on the pull request:
https://github.com/apache/spark/pull/2607#issuecomment-57554926
@manishamde Thanks for the WIP PR!
About classification, what points need to be discussed? Why is it more
difficult to figure out than regression? (Also, I personally am not a big fan
of the name "deviance" even though it is used in sklearn and in Friedman's
paper. I prefer more descriptive names like LogLoss.)
Also, will this be generalized to support weighted weak hypotheses, common
in most boosting algorithms?
For the final Model produced, should we use the same class for both random
forests and gradient boosting? It could be a TreeEnsemble model (to be
generalized later to a WeightedEnsemble model).
---
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 [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
