A generalizable dynamic flow pairing method for traffic classification.
José CamachoPablo PadillaPedro García-TeodoroJesús Esteban Díaz VerdejoPublished in: Comput. Networks (2013)
Keyphrases
- dimensionality reduction
- feature extraction
- feature selection
- classification accuracy
- classification method
- support vector machine svm
- support vector machine
- feature set
- preprocessing
- high precision
- pairwise
- cost function
- pattern classification
- high accuracy
- machine learning methods
- text classification
- benchmark data sets
- decision rules
- classification process
- probabilistic model
- computational cost
- significant improvement
- clustering method
- detection method
- support vector
- similarity measure
- decision trees
- classification rate
- extracted features
- classification procedure
- real time
- ordinal data
- network traffic
- cross validation
- segmentation method
- training samples
- dynamic programming
- computational complexity
- genetic algorithm
- neural network