Jon Machta, UMass, Amherst
“Population Annealing: A new algorithm for equilibrating spin glasses and other systems with rough free energy landscapes”

Systems with frustration and rough free energy landscapes such as spin glasses are very difficult to simulate using conventional Monte Carlo methods. I will describe a new approach introduced by Hukushima and Iba, population annealing, for sampling equilibrium states of systems with rough free energy landscapes. Population annealing is inspired by evolutionary dynamics. It is related to simulated annealing and is an example of a sequential Monte Carlo method rather than the more familiar Markov chain Monte Carlo method. I will show that population annealing is comparable in efficiency to state-of-the-art Markov chain Monte Carlo methods and has some advantages over these methods. Finally, I will present some new results for the three-dimensional Ising spin glass.