Team for Research in
Ubiquitous Secure Technology

Citation
"Evading Anomaly Detection through Variance Injection Attacks on PCA". B. Rubinstein, B. Nelson, L. Huang, A. Joseph, S. Lau, N. Taft, and J. D. Tygar (eds.), September, 2008.

Abstract
Whenever machine learning is applied to security problems, it is important to measure vulnerabilities to adversaries who poison the training data. We demonstrate the impact of variance injection schemes on PCA-based network-wide volume anomaly detectors, when a single compromised PoP injects chaff into the network. These schemes can increase the chance of evading detection by sixfold, for DoS attacks.

Electronic downloads

• http://www.cs.berkeley.edu/~tygar/papers/SML/RAID.08.pdf

• HTML

   <a
  href="http://www.truststc.org/pubs/729.html"
  ><i>Evading Anomaly Detection through Variance
  Injection Attacks on PCA</i></a>, B. Rubinstein,
  B. Nelson, L. Huang, A. Joseph, S. Lau, N. Taft, and J. D.
  Tygar (eds.), September, 2008.

• Plain text

   "Evading Anomaly Detection through Variance Injection
  Attacks on PCA". B. Rubinstein, B. Nelson, L. Huang, A.
  Joseph, S. Lau, N. Taft, and J. D. Tygar (eds.), September,
  2008.

• BibTeX

  @proceedings{RubinsteinNelsonHuangJosephLauTaftTygar08_EvadingAnomalyDetectionThroughVarianceInjectionAttacks,
      title = {Evading Anomaly Detection through Variance
                Injection Attacks on PCA},
      editor = {B. Rubinstein, B. Nelson, L. Huang, A. Joseph, S.
                Lau, N. Taft, and J. D. Tygar},
      month = {September},
      year = {2008},
      abstract = {Whenever machine learning is applied to security
                problems, it is important to measure
                vulnerabilities to adversaries who poison the
                training data. We demonstrate the impact of
                variance injection schemes on PCA-based
                network-wide volume anomaly detectors, when a
                single compromised PoP injects chaff into the
                network. These schemes can increase the chance of
                evading detection by sixfold, for DoS attacks.},
      URL = {http://www.truststc.org/pubs/729.html}
  }
  

Posted by Jessica Gamble on 7 Apr 2010.
For additional information, see the Publications FAQ or contact webmaster at www truststc org.