Processing math: 0%


Variational quantum algorithms for trace norms and their applications
Sheng-Jie Li,Jin-Min Liang,Shu-Qian Shen,Ming Li
Algorithm 3. The QVTN algorithm based on unitary decomposition
1. Inputs: the decomposition (19) of H, parameterized quantum circuits U({\boldsymbol{\theta }}) with initial parameters {\boldsymbol{\theta }}, and error tolerance ε.
2. For any 1\leqslant i\leqslant n, applying the Hadamard test in figure 4 to compute the value:
{f}_{i}({\boldsymbol{\theta }})=\displaystyle \frac{1}{2}\left(\mathrm{tr}({U}_{i}U({\boldsymbol{\theta }}))+\mathrm{tr}({U}^{\dagger }({\boldsymbol{\theta }}){U}_{i}^{\dagger })\right). (20)
3. Compute the value of loss function:
{l}_{3}({\boldsymbol{\theta }})=\sum _{i=1}^{n}{k}_{i}{f}_{i}({\boldsymbol{\theta }}). (21)
4. Perform optimization process to maximize {l}_{3}({\boldsymbol{\theta }}), update {\boldsymbol{\theta }}.
5. Repeat steps 2–4, until the loss function satisfies | {\rm{\Delta }}{l}_{3}({\boldsymbol{\theta }})| \leqslant \epsilon .