Motivated by previous work on moment varieties for Gaussian distributions and their
mixtures, we study moment varieties for two other statistically important
two-parameter distributions: the inverse Gaussian and gamma distributions. In
particular, we realize the moment varieties as determinantal varieties and find their
degrees and singularities. We also provide computational evidence for algebraic
identifiability of mixtures, and study the identifiability degree and Euclidean distance
degree.
Keywords
method of moments, moment varieties, identifiability,
determinantal varieties