Bayesian nonparametric estimation of Tsallis diversity indices under Gnedin-Pitman priors (1404.3441v2)

Abstract: Tsallis entropy is a generalized diversity index first derived in Patil and Taillie (1982) and then rediscovered in community ecology by Keylock (2005). Bayesian nonparametric estimation of Shannon entropy and Simpson's diversity under uniform and symmetric Dirichlet priors has been already advocated as an alternative to maximum likelihood estimation based on frequency counts, which is negatively biased in the undersampled regime. Here we present a fully general Bayesian nonparametric estimation of the whole class of Tsallis diversity indices under Gnedin-Pitman priors, a large family of random discrete distributions recently deeply investigated in posterior predictive species richness and discovery probability estimation. We provide both prior and posterior analysis. The results, illustrated through examples and an application to a real dataset, show the procedure is easily implementable, flexible and overcomes limitations of previous frequentist and Bayesian solutions.