In the last years, the field of quantum many-body systems has seen a sharp development, specifically in simulating interacting systems. However, a crucial challenge for current computational methods is the efficient numerical simulation of nonequilibrium real-time evolution due to the high-dimensionality of quantum systems. The state-of-the-art shows that artificial neural network encodings of quantum many-body wave function efficiently describe the time dynamics. In this thesis, an efficient machine learning-inspired approach has been employed to simulate the time dynamics. The neural-network quantum states are used to approximate the implicit midpoint rule method for solving the time-dependent Schrödinger equation. The proposed neural network architecture for the wave-function ansatz is the RBM (Restricted Boltzmann Machine), which has been shown to represent the ground state of various Hamiltonians with high accuracy. As a concrete example, this thesis studies the application of the transverse field Ising model on a one-dimensional lattice of different sizes, which exhibits an accuracy comparable to the stochastic configuration method. To deal with the high-dimensionality of a transverse-field Ising model on a lattice with periodic boundary conditions, the use of the GPU is employed in delivering results for larger lattice sizes. Observations show that the use of the GPU is critical for achieving results in large systems.     «

In the last years, the field of quantum many-body systems has seen a sharp development, specifically in simulating interacting systems. However, a crucial challenge for current computational methods is the efficient numerical simulation of nonequilibrium real-time evolution due to the high-dimensionality of quantum systems. The state-of-the-art shows that artificial neural network encodings of quantum many-body wave function efficiently describe the time dynamics. In this thesis, an efficient m...     »