MixedNUTS: Training-Free Accuracy-Robustness Balance via Nonlinearly Mixed Classifiers.
Yatong BaiMo ZhouVishal M. PatelSomayeh SojoudiPublished in: CoRR (2024)
Keyphrases
- training set
- test set
- training and testing data
- computational efficiency
- training set size
- discriminative classifiers
- high accuracy
- training data
- training examples
- classification performances
- fold cross validation
- highest accuracy
- feature selection
- higher classification accuracy
- roc curve
- training process
- training samples
- classification accuracy
- prediction accuracy
- confusion matrix
- class imbalanced
- individual classifiers
- training speed
- input space
- ensemble pruning
- ensemble learning
- linear classifiers
- classifier training
- computational cost
- support vector
- trained classifiers
- manually labeled training data
- classification rate
- generalization ability
- logistic regression
- error rate
- naive bayes
- decision trees