Github user codedeft commented on a diff in the pull request:
https://github.com/apache/spark/pull/2435#discussion_r18009800
--- Diff:
mllib/src/main/scala/org/apache/spark/mllib/tree/impl/DecisionTreeMetadata.scala
---
@@ -128,13 +139,34 @@ private[tree] object DecisionTreeMetadata {
}
}
+ // Set number of features to use per node (for random forests).
+ val _featureSubsetStrategy = featureSubsetStrategy match {
+ case "auto" => if (numTrees == 1) "all" else "sqrt"
+ case _ => featureSubsetStrategy
+ }
+ val numFeaturesPerNode: Int = _featureSubsetStrategy match {
+ case "all" => numFeatures
+ case "sqrt" => math.sqrt(numFeatures).ceil.toInt
+ case "log2" => math.max(1, (math.log(numFeatures) /
math.log(2)).ceil.toInt)
--- End diff --
R's randomForest defaults to 1/3 for regression.
Anecdotally, I've seen problems where a larger number of features work
better -- log2/sqrt might be a bit too severe, particularly when you don't have
many features to begin with.
---
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]
