Sitzungen | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

FM: Fall Meeting

FM 21: Quantum Computation: Algorithms

FM 21.2: Talk

Montag, 23. September 2019, 17:00–17:15, 2006

Quantum Generative Adversarial Networks for Learning and Loading Random Distributions — •Christa Zoufal^1,2, Aurélien Lucchi², and Stefan Woerner¹ — ¹IBM Research - Zurich, Rueschlikon, Switzerland — ²ETH, Zurich, Switzerland

Quantum algorithms have the potential to outperform their classical counterparts in a variety of tasks. The realization of the advantage often requires the ability to load classical data efficiently into quantum states. However, the best known methods for loading generic data into an n-qubit state require O(2ⁿ) gates. This scaling can easily predominate the complexity of a quantum algorithm and, thereby, impair potential quantum advantage.

Our work demonstrates that quantum Generative Adversarial Networks (qGANs) facilitate efficient loading of generic probability distributions into quantum states. More specifically, the qGAN scheme employs the interplay of a quantum channel, a variational quantum circuit, and a classical neural network to learn the probability distribution underlying given data samples and load it into the quantum channel. The resulting quantum channel loads the learned distribution with O(poly(n)) gates and can, thus, enable the exploitation of quantum advantage induced by quantum algorithms, such as Quantum Amplitude Estimation.

We implement the qGAN distribution learning and loading method with Qiskit and test it using a quantum simulation as well as actual quantum processors provided by the IBM Q Experience. Furthermore, we demonstrate the use of qGANs in a quantum finance application.