Table 2.

Three-node line A.

Number of Instances	Small (50)	Medium (200)	Large (1000)
ME	7.2 ×10−4	2.4 ×10−5	3.2 ×10−4
RMSE	0.0054	0.0017	0.0039
Best Fidelity	0.6840	0.8836	0.8872

The optimal control state for this network is $|1\rangle \otimes |1\rangle \otimes |1\rangle$, while the best solution that the machine learning protocol provides is $\mathrm{(0.1785}|0\rangle +0.9839|1\rangle )$  $\mathrm{(0.2063}|0\rangle +0.9785|1\rangle )$  $\mathrm{(0.1754}|0\rangle +0.9845|1\rangle )$.