Machine learning for adaptive quantum control and measurement

For many applications, like atomic clocks or gravitational wave detection, it is essential to precisely ascertain an optical phase. The goal of quantum measurement is to get as close to the fundamental Heisenberg limit as possible. I will present a method for designing measurement protocols that are closest to the Heisenberg limit. Our approach is based on a self-learning particle swarm algorithm that is trained on a simulated experiment to perform optimal phase estimation. Our algorithm learns solely based on training without any knowledge about the physical system. I will explain how this technique can be extended such that the machine learning algorithm can be trained on a real wold experiment.