A Nonlinear Regression for Characterizing Two-Qubit Processors
Melinda Andrews, Thomas Halverson, Joshua Heath, I. Michael Mandelberg, Martin J. McHugh, Shawn M. Wilder · 2023
We articulate a nonlinear regression to characterize the input state and measurement parameters of two-qubit computing systems. We represent state and (one-qubit product) measurement parameters and their constraints as products of Pauli and identity matrices. We apply our nonlinear regression on in-house simulated data. Our simulation models (i) nonideal state initialization, (ii) state dephasing and relaxation errors that occur during initialization, gate application, and measurement, (iii) gate control errors, and (iv) nonideal measurement. Our numerical regression returns physical, nontrivial state and measurement parameterizations with, respectively, high fidelity and small operator distance in our test cases. We are still experimenting with and refining this method and. here. present early findings.