Improved Threshold Selection by Using Calibrated Probabilities for Random Forest Classifiers.
Florian BaumannJinghui ChenKarsten VogtBodo RosenhahnPublished in: CRV (2015)
Keyphrases
- random forest
- threshold selection
- decision trees
- feature set
- fold cross validation
- rotation forest
- ensemble classifier
- ensemble learning
- random forests
- ensemble methods
- naive bayes
- feature ranking
- feature selection
- cancer classification
- majority voting
- decision tree learning algorithms
- machine learning algorithms
- feature vectors
- support vector
- training data
- base classifiers
- classification accuracy
- feature importance
- neural network
- classification models
- machine learning methods
- class labels
- test set
- training examples
- nearest neighbor classifiers