Harnessing deep neural networks to solve inverse problems in quantum dynamics: machine-learned predictions of time-dependent optimal control fields

Abstract

Inverse problems continue to garner immense interest in the physical sciences, particularly in the context of controlling desired phenomena in non-equilibrium systems. In this work, we utilize a series of...

Harnessing deep reinforcement learning to construct time-dependent optimal fields for quantum control dynamics

“…Similar to our previous work on quantum control of molecular systems, 23 we generated a set of potentials { V i ( x )} i =1,2,… of the form:where , , and , and denotes a uniform distribution. A comprehensive listing of the parameters used to generate the various potentials, V i ( x ), in this work can be found in the ESI †.…”

“…Similar to our previous work on quantum control of molecular systems, 23 we generated a set of potentials {V i (x)} i=1,2,. .…”

“…16,22 To address the previously mentioned computational bottlenecks, our group recently explored the use of supervised machine learning to solve these complex, inverse problems in quantum dynamics. 23 In contrast to supervised machine learning, reinforcement learning (RL) techniques have attracted recent attention since these machine learning methods are designed to solve sequential decision-making tasks, which can be naturally suited for quantum control problems. However, all prior RL studies to date have focused on low-dimensional spin-1/2 systems, which generally require a relatively small number of control pulses (typically 10-100) to converge.…”

Harnessing Deep Reinforcement Learning to Construct Time-Dependent Optimal Fields for Quantum Control Dynamics

“…Similar to our previous work on quantum control of molecular systems [23], we generated a set of potentials {V i (x)} i=1,2,... of the form:…”

“…To address the previously mentioned computational bottlenecks, our group recently explored the use of supervised machine learning to solve these complex, inverse problem in quantum dynamics [23]. In contrast to supervised machine learning, reinforcement learning (RL) techniques have attracted recent attention since these machine learning methods are designed to solve sequential decision-making tasks, which can be naturally suited for quantum control problems.…”