Classification and Clustering for Homeland Security Applications

This article introduces the basic concepts in classification and clustering and discusses their applications in homeland security. Both topics have been extensively studied in the fields of machine learning and data mining. As a result, many algorithms and heuristics have been developed to derive high-level knowledge from raw data. However, these algorithms are not magic black boxes that can automatically make decisions about complex problems and replace human beings. They require careful formulation of the real-world problem and tuning of parameters in order to be effective. While they might never replace a human being for making complex decisions, they are quite effective at modeling at well-formed problems. This article explores what are well-formed problems, what classification and clustering algorithms are capable of, and how can one use them effectively in the context of homeland security. Keywords: classification; clustering; supervised learning; unsupervised learning; machine learning