Active Learning Approach to Optimization of Experimental Control

Abstract

In this work we present a general machine learning based scheme to optimize experimental control. The method utilizes the neural network to learn the relation between the control parameters and the control goal, with which the optimal control parameters can be obtained. The main challenge of this approach is that the labeled data obtained from experiments are not abundant. The central idea of our scheme is to use the active learning to overcome this difficulty. As a demonstration example, we apply our method to control evaporative cooling experiments in cold atoms. We have first tested our method with simulated data and then applied our method to real experiments. We demonstrate that our method can successfully reach the best performance within hundreds of experimental runs. Our method does not require knowledge of the experimental system as a prior and is universal for experimental control in different systems.

References in corpus (26)

• Less is more: sampling chemical space with active learning
• Bose-Einstein Condensation of Atomic Hydrogen
• Active Learning of Uniformly Accurate Inter-atomic Potentials for Materials Simulation
• Bose-Einstein Condensation of Metastable Helium
• Reinforcement Learning in Different Phases of Quantum Control
• Fast machine-learning online optimization of ultra-cold-atom experiments
• Experimental neural network enhanced quantum tomography
• Fast and Accurate Uncertainty Estimation in Chemical Machine Learning
• Integrating Neural Networks with a Quantum Simulator for State Reconstruction
• When does reinforcement learning stand out in quantum control? A comparative study on state preparation
• Deep Reinforcement Learning for Quantum Gate Control
• Evaporative Cooling of a Two-Component Degenerate Fermi Gas
• Bose-Einstein Condensate in a light-induced vector potential using the 1064 $nm$ optical dipole trap lasers
• Multiparameter optimisation of a magneto-optical trap using deep learning
• Reinforcement learning for autonomous preparation of Floquet-engineered states: Inverting the quantum Kapitza oscillator
• Machine Learning for Achieving Bose-Einstein Condensation of Thulium Atoms
• Active emulation of computer codes with Gaussian processes -- Application to remote sensing
• Automatically growing global reactive neural network potential energy surfaces: a trajectory free active learning strategy
• Quantum Adiabatic Algorithm Design using Reinforcement Learning
• Applying machine learning optimization methods to the production of a quantum gas
• Approaching the adiabatic timescale with machine-learning
• Extreme Spin Squeezing from Deep Reinforcement Learning
• Retrieving Quantum Information with Active Learning
• Active Learning Algorithm for Computational Physics
• Dispersion Characterization and Pulse Prediction with Machine Learning
• A quantum active learning algorithm for sampling against adversarial attacks

Cited by in corpus (8)

• Faster State Preparation across Quantum Phase Transition Assisted by Reinforcement Learning
• Machine-learning-accelerated Bose-Einstein condensation
• Adaptive Quantum State Tomography with Active Learning
• High-dimensional reinforcement learning for optimization and control of ultracold quantum gases
• Review and experimental benchmarking of machine learning algorithms for efficient optimization of cold atom experiments
• Preparing Quantum States by Measurement-feedback Control with Bayesian Optimization
• Active Learning in Physics: From 101, to Progress, and Perspective
• Active Learning for Quantum Mechanical Measurements