Avoiding Boosting Overfitting by Removing Confusing Samples.

Alexander Vezhnevets Olga Barinova

Published in: ECML (2007)

Keyphrases

training error
data sets
ensemble methods
avoid overfitting
ensemble learning
decision trees
feature selection
training set
cross validation
real time
case study
error rate
machine learning methods
base classifiers
combining multiple
data samples
sample points
learning algorithm
unseen data