On the Use of Principle Component Analysis for the Hurst Parameter Estimation of Long-Range Dependent Network Traffic.
Melike ErolTayfun AkgülSema F. OktugSüleyman BaykutPublished in: ISCIS (2006)
Keyphrases
- long range
- parameter estimation
- network traffic
- principle component analysis
- independent component analysis
- markov random field
- intrusion detection
- anomaly detection
- maximum likelihood
- conditional random fields
- least squares
- svm classifier
- feature extraction
- dimension reduction
- intrusion detection system
- em algorithm
- model selection
- face recognition
- random fields
- feature space
- principal component analysis
- fractional brownian motion
- expectation maximization
- high dimensional
- probabilistic model
- pattern recognition
- covariance matrix
- higher order
- knowledge discovery
- image segmentation
- data sets