Penalized AdaBoost: Improving the Generalization Error of Gentle AdaBoost through a Margin Distribution.
Shuqiong WuHiroshi NagahashiPublished in: IEICE Trans. Inf. Syst. (2015)
Keyphrases
- generalization error
- minimum margin
- boosting algorithms
- training error
- learning algorithm
- weak learners
- model selection
- training data
- cross validation
- active learning
- loss function
- adaboost algorithm
- upper bound
- classification error
- binary classification
- weak classifiers
- sample complexity
- multi class
- training set
- sample size
- linear classifiers
- supervised learning
- risk minimization
- cost sensitive
- vc dimension
- face detection
- support vector
- object detection
- maximum margin
- target function
- error rate
- base classifiers
- conditional expectation
- worst case
- learning problems
- rademacher complexity
- neural network