Active Learning Approach to Optimization of Experimental Control
This work addresses the challenge of experimental control optimization in fields like physics, offering a universal method that is incremental in applying active learning to overcome data scarcity.
The authors tackled the problem of optimizing experimental control with limited labeled data by using an active learning approach that learns the relationship between control parameters and goals via a neural network. They demonstrated the method on evaporative cooling experiments with cold atoms, achieving the best performance within hundreds of runs without prior system knowledge.
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.