Hannover 2020 – wissenschaftliches Programm

Die DPG-Frühjahrstagung in Hannover musste abgesagt werden! Lesen Sie mehr ...

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

Q: Fachverband Quantenoptik und Photonik

Q 22: Posters: Quantum Optics and Photonics II

Q 22.39: Poster

Dienstag, 10. März 2020, 16:30–18:30, Empore Lichthof

Analysis of restricted Boltzmann machine quantum states — •Sebastian Syrkowski and Martin Gärttner — Kirchhoff-Institut für Physik, Heidelberg, Deutschland

For simulations of quantum many-body systems such as the transverse-field Ising model a major challenge one has to face is the exponentially growing complexity of the wave function with the number of particles prohibiting a simulation of larger systems on classical computers. Generative neural networks in the form of restricted Boltzmann machines (RBMs) have been shown to be able to represent such a quantum state more efficiently. In order to better understand the limits of this representation with respect to the complex optimization problems of ground state search and time evolution we are analyzing the topological structure of the underlying manifold of the RBM weight parameters using dimensionality reduction techniques. This allows us to test different optimization algorithms for their susceptibility to local minima.