[
https://issues.apache.org/jira/browse/MAHOUT-1419?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13905056#comment-13905056
]
Ted Dunning commented on MAHOUT-1419:
-------------------------------------
With t-digest in the OnlineSummarizer now, it is quite possible to have very
accurate and quick quantile estimates. That should allow very quick picking of
splits as well.
The basic idea would be to keep an OnlineSummarizer for each numerical
variable. When a split is needed, pick a random number in [0,1) and then the
split point is that quantile. If you want 10 random splits, do this 10 times.
If you want structured points like every percentile from 20 to 80 %, that is
just as simple.
It seems like this would be as simple as changing a loop from looping over
distinct values to looping over quantile values.
> Random decision forest is excessively slow on numeric features
> --------------------------------------------------------------
>
> Key: MAHOUT-1419
> URL: https://issues.apache.org/jira/browse/MAHOUT-1419
> Project: Mahout
> Issue Type: Bug
> Components: Classification
> Affects Versions: 0.7, 0.8, 0.9
> Reporter: Sean Owen
>
> Follow-up to MAHOUT-1417. There's a customer running this and observing it
> take an unreasonably long time on about 2GB of data -- like, >24 hours when
> other RDF M/R implementations take 9 minutes. The difference is big enough to
> probably be considered a defect. MAHOUT-1417 got that down to about 5 hours.
> I am trying to further improve it.
> One key issue seems to be how splits are evaluated over numeric features. A
> split is tried for every distinct numeric value of the feature in the whole
> data set. Since these are floating point values, they could (and in the
> customer's case are) all distinct. 200K rows means 200K splits to evaluate
> every time a node is built on the feature.
> A better approach is to sample percentiles out of the feature and evaluate
> only those as splits. Really doing that efficiently would require a lot of
> rewrite. However, there are some modest changes possible which get some of
> the benefit, and appear to make it run about 3x faster. That is --on a data
> set that exhibits this problem -- meaning one using numeric features which
> are generally distinct. Which is not exotic.
> There are comparable but different problems with handling of categorical
> features, but that's for a different patch.
> I have a patch, but it changes behavior to some extent since it is evaluating
> only a sample of splits instead of every single possible one. In particular
> it makes the output of "OptIgSplit" no longer match the "DefaultIgSplit".
> Although I think the point is that "optimized" may mean giving different
> choices of split here, which could yield differing trees. So that test
> probably has to go.
> (Along the way I found a number of micro-optimizations in this part of the
> code that added up to maybe a 3% speedup. And fixed an NPE too.)
> I will propose a patch shortly with all of this for thoughts.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
