A Method of Detecting the Abnormal Encrypted Traffic Based on Machine Learning and Behavior Characteristics.
Bin KongZhangpu LiuGuangming ZhouXiaoyan YuPublished in: ICCNS (2019)
Keyphrases
- machine learning
- significant improvement
- dynamic programming
- objective function
- machine learning methods
- prior knowledge
- computational cost
- high accuracy
- cost function
- synthetic data
- optimization algorithm
- similarity measure
- preprocessing
- support vector machine
- classification accuracy
- high precision
- statistical methods
- fully automatic
- real time
- edge detection
- experimental evaluation
- active learning
- pairwise
- pattern recognition
- neural network