One-Class SVM with Privileged Information and its Application to Malware\n Detection

2016/09/26 by Evgeny Burnaev, Burnaev, Evgeny, Dmitry Smolyakov +1
Computer Science · Engineering · #Advanced Malware Detection Techniques #Anomaly Detection Techniques and Applications #Applications (stat.AP) #Artificial Immune Systems Applications #Cryptography and Security (cs.CR) #FOS: Computer and information sciences #Machine Learning (stat.ML)

paper · pdf · doi:10.48550/arxiv.1609.08039

openalex publication_date 2016/09/26 · openalex created_date 2022/09/25 · openalex updated_date 2026/07/28

Abstract

A number of important applied problems in engineering, finance and medicine\ncan be formulated as a problem of anomaly detection. A classical approach to\nthe problem is to describe a normal state using a one-class support vector\nmachine. Then to detect anomalies we quantify a distance from a new observation\nto the constructed description of the normal class. In this paper we present a\nnew approach to the one-class classification. We formulate a new problem\nstatement and a corresponding algorithm that allow taking into account a\nprivileged information during the training phase. We evaluate performance of\nthe proposed approach using a synthetic dataset, as well as the publicly\navailable Microsoft Malware Classification Challenge dataset.\n

Related

• Anomaly Detection Under Uncertainty Using Distributionally Robust Optimization Approach
• Calibration of One-Class SVM for MV set estimation
• Anomaly Classification with the Anti-Profile Support Vector Machine
• What is Wrong with One-Class Anomaly Detection?
• Security Evaluation of Support Vector Machines in Adversarial Environments