Abstract

We construct an algorithm that learns through trial-and-error training how to devise optimal feedback-based single-shot phase estimation in interferometry. Our algorithm is robust against experimental imperfections, losses and decoherence.

© 2012 OSA

PDF Article