Figure 1

(a) In quantum device modeling, we guess a model that is consistent with the quantum theory and the statistics of data from the experiment. However, such a model may not reflect actual physics in the system. (b) If there is only one model consistent with the corresponding probability distribution, then the model can always be correctly identified. Here, we assume that the statistical fluctuation is negligible provided sufficient data; i.e., we ignore the error in the estimation of the probability distribution. (c) If there is more than one model consistent with the corresponding probability distribution, the model $[{\mathfrak{m}}^{\prime }]$ reconstructed from the distribution may be different from the genuine model of the system $[\mathfrak{m}]$ that causes the experimental result.