Fig. 1
From: Post-variational classical quantum transfer learning for binary classification
This is a high level end to end flow of the proposed PVCQTL model. The regular parameterized ansatz is replaced by Post Variational Design Strategies: (1) The modified observable construction approach uses a classical combination of predefined trial observables. (2) The hybrid approach uses an ansatz with a single repetition (for reduced depth) and classically combines multiple observables. In this approach, the parameters are not tuned by an optimizer but are predefined by the parameter shift rule. (3) In the variational with post-variational approach, a parameterized ansatz is combined with a modified observable construction. Here, the variational parameters are tuned by an optimizer according to a cost function, and multiple observables are measured, as in the modified observable construction approach.
