Fig. 2: Quantum process tomography with tensor networks.
From: Quantum process tomography with unsupervised learning and tensor networks
a The quantum process (N = 4) is represented by a Choi matrix Λϑ, parametrized by a locally-purified density operator (LPDO). The input and output indices of the process are {σj} and {τj}, respectively. b Tensor contraction evaluating the conditional probability distribution Pϑ(β∣α), i.e., the probability that the LPDO Choi matrix associates with the measurement Mβ given the state \({{{{{{{{\boldsymbol{\rho }}}}}}}}}_{{{{{{{{\boldsymbol{\alpha }}}}}}}}}={t}_{{{{{{{{\boldsymbol{\alpha }}}}}}}}}^{-1}{{{{{{{{\boldsymbol{M}}}}}}}}}_{{{{{{{{\boldsymbol{\alpha }}}}}}}}}\) at the input of the channel.
