On the Effectiveness of Activation Noise in Both Training and Inference for Generative Classifiers.
Milad Khademi NoriYiqun GeIl-Min KimPublished in: IEEE Access (2023)
Keyphrases
- training set
- discriminative classifiers
- training and testing data
- training examples
- training samples
- test set
- training data
- discriminative training
- noise level
- svm classifier
- feature selection
- classifier training
- supervised learning
- information processing
- support vector
- labeled training data
- decision trees
- structured prediction
- training process
- multi layer perceptron
- data sets
- linear svm
- boosted classifiers
- text classifiers
- probabilistic inference
- class labels
- missing data
- semi supervised learning
- labeled data
- unsupervised learning
- generative model
- active learning