Bootstrap aggregating
Agregación de bootstrap
- Date
- 1994-1996 · confidence low
- Place
- Berkeley, California, United States
- Credited to
- Leo Breiman
- Field
- Knowledge & symbols
What it is
Machine learning meta-algorithm devised by Leo Breiman that combines several models trained on bootstrap-resampled subsets of the same dataset to improve prediction stability and accuracy.
What it led to
- Reduced variance and overfitting in decision trees
- Direct basis for random forests
- Widespread adoption as an ensemble technique in machine learning