When characterizing quantum systems, quantum process tomography (QPT) is the standard primitive. But due to the high complexity of quantum systems and the curse of dimensionality, QPT becomes impractical when dealing with a large number of qubits. On the other hand, combining QPT and machine learning has shown great success in recent studies. In this thesis, the opportunity is explored of doing QPT in combination with machine learning and parametrized quantum circuits, regarding the reconstruction of Hamiltonians of spin glasses. This results in a rather simple and straightforward algorithm. For this, in the beginning, the necessary quantum circuit is derived. With this, the Hamiltonians of Ising spins are reconstructed. Finally, we switch to spin glasses, which doesn’t differ much to Ising spins, and do the same here. From this, the systems are fully characterized by the obtained Hamiltonians afterwards. These approaches are done for system sizes of up to 12 qubits, whereas more qubits would be also possible. The results of the reconstructions are reaching high fidelity values using simulated data for the Ising model and spin glasses, showing and underlining the efficiency of the proposed algorithm.     «

When characterizing quantum systems, quantum process tomography (QPT) is the standard primitive. But due to the high complexity of quantum systems and the curse of dimensionality, QPT becomes impractical when dealing with a large number of qubits. On the other hand, combining QPT and machine learning has shown great success in recent studies. In this thesis, the opportunity is explored of doing QPT in combination with machine learning and parametrized quantum circuits, regarding the reconstructi...     »