Frequentist Parameter Estimation with Supervised Learning
arXiv:2105.12302 · doi:10.1116/5.0058163
Abstract
Recently there has been a great deal of interest surrounding the calibration of quantum sensors using machine learning techniques. In this work, we explore the use of regression to infer a machine-learned point estimate of an unknown parameter. Although the analysis is neccessarily frequentist - relying on repeated esitmates to build up statistics - we clarify that this machine-learned estimator converges to the Bayesian maximum a-posterori estimator (subject to some regularity conditions). When the number of training measurements are large, this is identical to the well-known maximum-likelihood estimator (MLE), and using this fact, we argue that the Cram{é}r-Rao sensitivity bound applies to the mean-square error cost function and can therefore be used to select optimal model and training parameters. We show that the machine-learned estimator inherits the desirable asymptotic properties of the MLE, up to a limit imposed by the resolution of the training grid. Furthermore, we investigate the role of quantum noise the training process, and show that this noise imposes a fundamental limit on number of grid points. This manuscript paves the way for machine-learning to assist the calibration of quantum sensors, thereby allowing maximum-likelihood inference to play a more prominent role in the design and operation of the next generation of ultra-precise sensors.
References in corpus (34)
- Quantum sensing
- Machine learning and the physical sciences
- Quantum metrology with nonclassical states of atomic ensembles
- A high-bias, low-variance introduction to Machine Learning for physicists
- Entanglement, Non-linear Dynamics, and the Heisenberg Limit
- Twin matter waves for interferometry beyond the classical limit
- Entanglement-enhanced measurement of a completely unknown phase
- Machine Learning for Precise Quantum Measurement
- Optimized quantum sensing with a single electron spin using real-time adaptive measurements
- Phase detection at the quantum limit with multi-photon Mach-Zehnder interferometry
- Useful Multiparticle Entanglement and Sub-Shot-Noise Sensitivity in Experimental Phase Estimation
- An Efficient Algorithm for Optimizing Adaptive Quantum Metrology Processes
- Approaching quantum-limited metrology with imperfect detectors by using weak-value amplification
- Generalizable control for quantum parameter estimation through reinforcement learning
- Experimental Phase Estimation Enhanced By Machine Learning
- Weak-value amplification and optimal parameter estimation in the presence of correlated noise
- Differential Evolution for Many-Particle Adaptive Quantum Metrology
- Bayesian estimation in homodyne interferometry
- Calibration of quantum sensors by neural networks
- Frequentist and Bayesian Quantum Phase Estimation
- A differentiable programming method for quantum control
- Improving the dynamics of quantum sensors with reinforcement learning
- Neural-Network Heuristics for Adaptive Bayesian Quantum Estimation
- A Machine-Designed Sensor to Make Optimal Use of Entanglement-Generating Dynamics for Quantum Sensing
- Magnetic-field-learning using a single electronic spin in diamond with one-photon-readout at room temperature
- Robust calibration of multiparameter sensors via machine learning at the single-photon level
- Non-asymptotic analysis of quantum metrology protocols beyond the Cramér-Rao bound
- Machine-learning-assisted electron-spin readout of nitrogen-vacancy center in diamond
- NV center based nano-NMR enhanced by deep learning
- Robustness of Quantum-Enhanced Adaptive Phase Estimation
- Feedback Ansatz for Adaptive-Feedback Quantum Metrology Training with Machine Learning
- Time-dependent atomic magnetometry with a recurrent neural network
- Quantum parameter estimation with a neural network
- Diagnosing Imperfections in Quantum Sensors via Generalized Cramér-Rao Bounds