Signal/noise optimization strategies for stochastically estimated correlation functions

Abstract

Numerical studies of quantum field theories usually rely upon an accurate determination of stochastically estimated correlation functions in order to extract information about the spectrum of the theory and matrix elements of operators. The reliable determination of such correlators is often hampered by an exponential degradation of signal/noise at late time separations. We demonstrate that it is sometimes possible to achieve significant enhancements of signal/noise by appropriately optimizing correlators with respect to the source and sink interpolating operators, and highlight the large range of possibilities that are available for this task. The ideas are discussed for both a toy model, and single hadron correlators in the context of quantum chromodynamics.