Recent Advances in Intrusion Detection

Volume 5230 of the series Lecture Notes in Computer Science pp 394-395

Evading Anomaly Detection through Variance Injection Attacks on PCA

(Extended Abstract)
  • Benjamin I. P. RubinsteinAffiliated withUC Berkeley
  • , Blaine NelsonAffiliated withUC Berkeley
  • , Ling HuangAffiliated withIntel Research
  • , Anthony D. JosephAffiliated withUC BerkeleyIntel Research
  • , Shing-hon LauAffiliated withUC Berkeley
  • , Nina TaftAffiliated withIntel Research
  • , J. D. TygarAffiliated withUC Berkeley

* Final gross prices may vary according to local VAT.

Get Access


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.