Density of state method for complex action systems

Abstract

While importance sampling Monte Carlo algorithms have proved to be a crucial tool for numerical studies in modern physics, they fail when we consider complex action systems. The density of states approach provides a way to simulate such systems and reduce the sign problem that afflicts them to a 1-dimensional oscillatory integral. In this work, we shall review the density of states approach as well as the Linear Logarithmic Relaxation algorithm and present some recent development concerning the control of systematics in this algorithm. The results of a benchmark study on the relativistic Bose gas shall be presented as well.