One-Class Strategies for Security Information Detection

摘要

Detecting security-related information is a critical component of ISI research, which involves studying a wide range of technical and systems challenges related to the acquisition, collection, storage, retrieval, synthesis, and analysis of security-related information. Outlier or anomaly detection is a well-known problem in statistics which can be naturally described as Density Level Detection problems and outliers are intuitively understood as events with small probability. Unfortunately, these algorithms cannot be directly used for general security-related information detection problems since the imposed assumptions on the densities are often tailored to specific applications. Instead, the well-known Statistical Learning Theory (SLT) provides distribution-free conditions and guarantees for good performance of generalization for learning algorithms. In statistical machine learning, outlier or anomaly detection can be equivalently described as one-class problems.