Green's function monte carlo and the many-fermion problem

Abstract

The application of Green's function Monte Carlo to many body problems is outlined. For boson problems, the method is well developed and practical. An “efficiency principle”,importance sampling, can be used to reduce variance. Fermion problems are more difficult because spatially antisymmetric functions must be represented as a difference of two density functions. Naively treated, this leads to a rapid growth of Monte Carlo error. Methods for overcoming the difficulty are discussed. Satisfactory algorithms exist for few-body problems; for many-body problems more work is needed, but it is likely that adequate methods will soon be available.KeywordsImportance SamplingTrial FunctionNodal SurfaceMonte Carlo ErrorAntisymmetric FunctionThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.