Pareto-optimal quantum kernel selection for unsupervised anomaly detection on real malware beaconing data
Pareto quantum-kernel selection on real malware beaconing slightly beats a classical one-class SVM baseline.
The authors propose an unsupervised multi-objective protocol that jointly optimizes normalised pseudo discrepancy and geometric difference to a classical kernel, then selects quantum kernels from the Pareto front. Applied to malware beaconing detection with a one-class SVM, fidelity and projected quantum kernels are tested over four encodings on simulators and IQM's 20-qubit Garnet processor. An NPD-selected fidelity kernel reaches AUC 0.840 versus a tuned classical baseline of 0.765, but its geometric difference is too small to certify a quantum gain. The Pareto-selected projected kernel only marginally exceeds the baseline at AUC 0.782 despite a large geometric difference.
- Protocol jointly optimizes detection quality and geometric difference.
- Fidelity kernel AUC 0.840 beats a classical baseline of 0.765.
- That gain is not certified as quantum because geometric difference is small.
- Pareto projected kernel reaches AUC 0.782 on IQM's 20-qubit Garnet.
From the original article
1 min read at arxiv.orgQuantum kernel methods are leading candidates for a practical quantum advantage in machine learning, but assessing that potential requires two quantities usually reported separately: how well a kernel performs on the task, and how far its geometry departs from the classical kernels available for the same problem.
By Boaz Micah · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.