1304.5828 (Christopher Ferrie et al.)

Likelihood-free quantum inference: tomography without the Born Rule    [PDF]

Christopher Ferrie, Christopher E. Granade

We introduce the likelihood-free quantum inference algorithm (LFQIA), a Bayesian learning algorithm which can estimate quantum states and processes given access to a classical experiment simulator rather than to exact likelihood functions. It has been shown recently that there exists interesting classes of states and processes for which weak simulation (the ability to efficiently sample the probability distribution of outcomes of the experiment) is possible but strong simulation (the ability to efficiently compute the full distribution of outcomes of the experiment) is not. In such cases, our algorithm exponentially outperforms standard algorithms based on computation of the likelihood function and we demonstrate this with an example.

View original: http://arxiv.org/abs/1304.5828