Randomized ensemble methods for classification trees
Approved for public release, distribution is unlimited === We propose two methods of constructing ensembles of classifiers. One method directly injects randomness into classification tree algorithms by choosing a split randomly at each node with probabilities proportional to the measure of goodness...
| Main Author: | |
|---|---|
| Other Authors: | |
| Published: |
Monterey, California. Naval Postgraduate School
2012
|
| Online Access: | http://hdl.handle.net/10945/9801 |