Development of a Suitable Framework for Empirical Evaluation of Pattern Classification Systems to Improve Performance and Enhance Security

Abstract

In this research paper, we address one of the main open issues: evaluating at design phase the security of pattern classifiers, namely, the performance degradation under potential attacks they may incur during operation. We propose a framework for empirical evaluation of classifier security that formalizes and generalizes the main ideas proposed in the literature, and give examples of its use in three real applications. Reported results show that security evaluation can provide a more complete understanding of the classifier’s behaviour in adversarial environments and lead to better design choices.
KEYWORDS: Classification and Regression Trees (CART); Chi Square Automatic Interaction Detection (CHAID); Intrusion Detection Systems (IDS); Polymorphic Blending Attacks; Data Visualization