Listen to Minority: Encrypted Traffic Classification for Class Imbalance with Contrastive Pre-Training.
Xiang LiJuncheng GuoQige SongJiang XieYafei SangShuyuan ZhaoYongzheng ZhangPublished in: SECON (2023)
Keyphrases
- class imbalance
- class imbalanced
- class distribution
- majority class
- minority class
- training set
- cost sensitive learning
- cost sensitive
- imbalanced datasets
- training samples
- active learning
- supervised learning
- high dimensionality
- test set
- misclassification costs
- training examples
- imbalanced class distribution
- training data
- imbalanced data sets
- imbalanced data
- feature space
- sampling methods
- concept drift
- training process
- classification error
- rare class
- nearest neighbour
- data sets
- class labels
- classification accuracy
- decision trees
- high dimensional
- machine learning methods
- attribute selection
- decision boundary
- training dataset
- generalization ability
- classification algorithm
- text categorization
- unlabeled data
- unsupervised learning
- dimensionality reduction
- support vector machine
- feature vectors
- support vector
- feature selection
- learning algorithm