Researchers from Multiverse Computing and CounterCraft have collaborated to develop a quantum-inspired tensor network based algorithm enhancing cyber attack detection. Known as the Matrix Product State (MPS) model, it achieved 100% detection of attacks, employing adversary-generated threat intelligence instead of rule-based systems. This innovation significantly reduces false positives compared to classical models while enhancing interpretability, crucial for business users and regulators. The model’s efficacy was tested on various attack types, demonstrating its ability to distinguish normal from abnormal behavior. This development marks a significant advancement in cybersecurity defense, leveraging quantum techniques to anticipate and counteract evolving threats. A press release describing this cybersecurity detection algorithm is available here and further technical details about this approach can be found in the paper posted on ArXiv: “Tensor Networks for Explainable Machine Learning in Cybersecurity.“
March 26, 2024