A class-oriented feature selection approach for multi-class imbalanced network traffic datasets based on local and global metrics fusion.
Zhen LiuRuoyu WangMing TaoXianfa CaiPublished in: Neurocomputing (2015)
Keyphrases
- network traffic
- class imbalanced
- feature selection
- instance selection
- class imbalance
- intrusion detection
- network traffic data
- anomaly detection
- intrusion detection system
- classification accuracy
- feature space
- multi class
- nearest neighbor
- imbalanced data
- high dimensionality
- minority class
- machine learning
- text classification
- dimensionality reduction
- unsupervised learning
- feature extraction
- long range dependence
- knn
- evaluation criteria
- cost sensitive
- support vector machine
- class distribution
- data warehouse
- multiple instance learning
- knowledge discovery and data mining
- support vector
- training data
- supervised learning
- learning algorithm
- normal traffic
- real world
- text categorization