Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

QI: Fachverband Quanteninformation

QI 10: Quantum Machine Learning II

QI 10.5: Vortrag

Dienstag, 11. März 2025, 12:00–12:15, HS VIII

Quantum Kernel Methods under Scrutiny — •Jan Schnabel, Roberto Flórez Ablan, and Marco Roth — Fraunhofer Institute for Manufacturing Engineering and Automation IPA, Stuttgart, Germany

Quantum kernel methods (QKMs) have emerged as a promising approach in quantum machine learning, offering both practical applications and theoretical insights. Two primary strategies for computing the Gram matrix in QKMs are fidelity quantum kernels (FQKs) and projected quantum kernels (PQKs). Benchmarking these methods is crucial to gain robust insights and to understand their practical utility.

In this talk, we present a comprehensive large-scale study examining QKMs based on FQKs and PQKs across a manifold of design choices, covering both classification and regression tasks. Our work spans five dataset families and 64 datasets, resulting in over 20,000 models trained and optimized using a state-of-the-art hyperparameter search. We delve into the importance of hyperparameters on model performance scores and provide a thorough analysis addressing the design freedom of PQKs and explore the underlying principles responsible for learning. Rather than pinpointing the best-performing models for specific tasks, our goal is to uncover the mechanisms that drive effective QKMs and reveal universal patterns. These insights contribute to better understand certain properties of QKMs and what distinguishes good from bad models.

Keywords: quantum kernel methods; quantum machine learning; benchmarking