Contracted Quantum Learning of Random Unitary Channels

Online learning algorithms for quantum state tomography, where the learning model is sequentially updated in response to new data, comprise a promising and exciting topic in contemporary quantum information research. Seeking to apply online learning to quantum process tomography, we give a protocol for analyzing a class of orthogonal random unitary channels using Pauli learning with a contracted quantum learning procedure. Our approach utilizes multi-objective Pauli and unitary minimization, and allows for the learning of locally equivalent channels. We demonstrate the success of this technique for varying degrees of noise, as well as characterize the scalability of this approach, particularly for sparse evaluations of operators.