Summary
This research article presents Valkyrie, a classification system designed to detect malicious binaries based on behavioral traits from large-scale telemetry data. Utilizing Apache Spark for efficient processing, Valkyrie achieves high confidence predictions with minimal endpoint load, making it suitable for production systems. The paper discusses the methodology, data collection, and the advantages of using behavioral analysis over traditional static methods in malware detection.
