| Title | Multiple Classifier Systems for Adversarial Classification Tasks |
| Publication Type | Conference Paper |
| Year of Publication | 2009 |
| Authors | Biggio, B, Fumera, G, Roli, F |
| Editor | Benediktsson, JAtli, Kittler, J, Roli, F |
| Conference Name | 8th Int. Workshop on Multiple Classifier Systems (MCS 2009) |
| Volume | 5519 |
| Pagination | 132-141 |
| Date Published | 10/06/2009 |
| Publisher | Springer |
| Conference Location | Reykjavik, Iceland |
| ISBN Number | 978-3-642-02325-5 |
| Keywords | adversarial classification, adversarial learning, mcs00, Multiple Classifier Systems |
| Abstract | Pattern classification systems are currently used in security applications like intrusion detection in computer networks, spam filtering and biometric identity recognition. These are adversarial classification problems, since the classifier faces an intelligent adversary who adaptively modifies patterns (e.g., spam e-mails) to evade it. In these tasks the goal of a classifier is to attain both a high classification accuracy and a high hardness of evasion, but this issue has not been deeply investigated yet in the literature. We address it under the viewpoint of the choice of the architecture of a multiple classifier system. We propose a measure of the hardness of evasion of a classifier architecture, and give an analytical evaluation and comparison of an individual classifier and a classifier ensemble architecture. We finally report an experimental evaluation on a spam filtering task.
|
| Notes |
|
| URL | http://www.springerlink.com/content/m851267062872650/ |
| Citation Key | 779 |
biggio09-mcs.pdf
