Boosting the margin: A new explanation for the effectiveness of voting methods

@inproceedings{Schapire1997BoostingTM,
  title={Boosting the margin: A new explanation for the effectiveness of voting methods},
  author={R. Schapire and Y. Freund and Peter Barlett and Wee Sun Lee},
  booktitle={ICML},
  year={1997}
}
One of the surprising recurring phenomena observed in experiments with boosting is that the test error of the generated classifier usually does not increase as its size becomes very large, and often is observed to decrease even after the training error reaches zero. In this paper, we show that this phenomenon is related to the distribution of margins of the training examples with respect to the generated voting classification rule, where the margin of an example is simply the difference between… Expand
Further results on the margin explanation of boosting: new algorithm and experiments
How boosting the margin can also boost classifier complexity
On the Insufficiency of the Large Margins Theory in Explaining the Performance of Ensemble Methods
Maximizing the Margin with Boosting
Further results on the margin distribution
Analyzing Margins in Boosting
On the Margin Explanation of Boosting Algorithms
Supervised projection approach for boosting classifiers
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 51 REFERENCES
Boosting in the Limit: Maximizing the Margin of Learned Ensembles
Experiments with a New Boosting Algorithm
An Empirical Evaluation of Bagging and Boosting
Bagging, Boosting, and C4.5
Structural Risk Minimization Over Data-Dependent Hierarchies
Improved Boosting Algorithms using Confidence-Rated Predictions
For Valid Generalization the Size of the Weights is More Important than the Size of the Network
Boosting Decision Trees
A training algorithm for optimal margin classifiers
...
1
2
3
4
5
...