Model-aware reinforcement learning for high-performance Bayesian experimental design in quantum metrology
Belliardo F, Zoratti F, Marquardt F, Giovannetti V (2024)
Publication Language: English
Publication Type: Journal article
Publication year: 2024
Journal
Book Volume: 8
Article Number: 1555
DOI: 10.22331/q-2024-12-10-1555
Abstract
Quantum sensors offer flexibility in control during estimation, allowing manipulation across various parameters by the experimenter. For each sensing platform, determining the optimal controls to enhance the sensor’s precision remains a challenging task. While an analytical solution may be unattainable, machine learning presents a promising approach for many systems of interest, especially considering the capabilities of modern hardware. We introduce a versatile procedure capable of optimizing a wide range of problems in quantum metrology and estimation by combining model-aware reinforcement learning (RL) with Bayesian estimation via particle filtering. To achieve this, we addressed the challenge of integrating the many non-differentiable steps of the estimation process, such as measurements and particle filter resampling, into the training routine. Our RL-based approach is suitable for optimizing both non-adaptive and adaptive strategies using a neural network. We provide an implementation of this technique in the form of a Python library called qsensoropt, along with several pre-built applications for relevant physical platforms, including NV centers, photonic circuits, and optical cavities. Using our method, we have achieved results that surpass the current state-of-the-art in experimental design for numerous tasks. Beyond Bayesian estimation, by leveraging model-aware RL, it is also possible to find optimal controls for minimizing the Cramér-Rao bound, based on Fisher information.
Authors with CRIS profile
Involved external institutions
Italy (IT)How to cite
APA:
Belliardo, F., Zoratti, F., Marquardt, F., & Giovannetti, V. (2024). Model-aware reinforcement learning for high-performance Bayesian experimental design in quantum metrology. Quantum, 8. https://doi.org/10.22331/q-2024-12-10-1555
MLA:
Belliardo, Federico, et al. "Model-aware reinforcement learning for high-performance Bayesian experimental design in quantum metrology." Quantum 8 (2024).
BibTeX: Download