Learning Terrorist Profiles by Optimization with Two Objectives

摘要

When profiling techniques are used in practice, in addition to identifying all the positive objects, it is very often that we want not only to minimize the total number of false alarms, but also use as few attributes as possible. In this paper, we propose an optimization-based approach to this problem. The proposed method could be used as a component in a machine learning system to extract profiling patterns for subsequent object classification.