A0265
Title: On the Voigt distribution: Characterization and parameter estimation
Authors: Gavino Puggioni - University of Rhode Island (United States) [presenting]
Massimo Cannas - University of Cagliari (Italy)
Abstract: The Voigt profile is the convolution of Gaussian and Cauchy random variables. The Voigt is extensively used in atomic and molecular spectroscopy to represent superposition effects. The lack of a moment-generating function and of a closed form for the density has generated some interest in the literature about parameter estimation. We provide a new characterization of the Voigt profile and its associated dual. We also propose an MCMC algorithm to estimate the posterior distribution of both scale and location parameters. A simulation study demonstrates a better performance of our algorithm compared to other approaches.
