Transcription of 1 RANDOM FORESTS
{{id}} {{{paragraph}}}
Example: air traffic controller
1 RANDOM FORESTS
1 RANDOM FORESTS Leo BreimanStatistics Department University of California Berkeley, CA 94720 January 2001 AbstractRandom FORESTS are a combination of tree predictorssuch that each tree depends on the values of a randomvector sampled independently and with the samedistribution for all trees in the forest. Thegeneralization error for FORESTS converges to a limitas the number of trees in the forest becomes generalization error of a forest of tree classifiersdepends on the strength of the individual trees in theforest and the correlation between them. Using arandom selection of features to split each node yieldserror rates that compare favorably to Adaboost(Freund and Schapire[1996]), but are more robust withrespect to noise.
An early example is bagging (Breiman [1996]), where to grow each tree a random selection (without replacement) is ... estimates of variable importance and binding these together by reuse runs. Section 11 looks at random forests for regression. A bound for the mean
Loading..
Tags:
Information
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document: