Cybersecurity Anomaly and Outlier Detection Validation
Abstract
The U.S. Army is teamed with innovative industry partners and has developed a quantum-inspired and quantum-ready and accelerated AI computing environment (Accelerated AI Platform) that offers speed, scale and accuracy for optimization and AI problems. The platforms analytics capabilities will assist the Government in quickly optimizing and accelerating AI and machine learning processes using Entanglements solver NGQ (trademark) (quantum optimization), and Neural Network applications for Cyber Threat Detection, specifically real-time anomaly detection. AAG seeks to apply NGQ(trademark) and novel AI hardware processors to three areas that would support the continuous monitoring portion of Zero Trust architectures. This includes an anomaly detection algorithm capable of continuously vetting all users on a network and their actions. A similar algorithmic framework will be suitable for demonstrating Intrusion Detection Systems (IDS) and expanded threat awareness at network endpoints to improve the processing of telemetry data dramatically within current cyber operations. In demonstrating such a capability, this work will have engineered a new class of anomaly detection algorithms capable of use not only for cyber, but for many problems where the events of interest happen very infrequently but are of great significance when they do.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 17, 2022
- Accession Number
- AD1180411
Entities
People
- Clay Stanek