Quantum kernel methods for anomaly detection in manifold-structured telemetry. Benchmarks entanglement-based kernels and classical baselines with geometry diagnostics to study low-data anomaly separation in non-Euclidean dynamical systems.

Quoptuna is a Python package supporting over 20 quantum machine learning models using PennyLane. These models include classifiers, neural networks, and kernel-based approaches. Quoptuna integrates seamlessly with quantum simulators, enabling model evaluation without requiring quantum hardware.