Benign Overfitting in Linear Classifiers and Leaky ReLU Networks from KKT Conditions for Margin Maximization.
Spencer FreiGal VardiPeter L. BartlettNathan SrebroPublished in: CoRR (2023)
Keyphrases
- linear classifiers
- margin maximization
- generalization error
- hyperplane
- multi class
- cross validation
- support vector
- hyperparameters
- active learning
- upper bound
- decision trees
- computer vision
- multi task
- model selection
- svm classifier
- metric learning
- image classification
- incremental learning
- principal components
- neural network
- support vector machine
- feature space
- feature extraction
- feature selection
- learning algorithm