Distribuição beta prime inflacionada em zero

Abstract

The probability distributions whether discrete or continuous can encounter barriers when we talk about data that contains "zeros in excess" LORD et al. (2005). In order for these distributions to measure such data, mixed distributions called inflated distributions are created. The way in which this inflation will occur will depend of the distribution support set . In this research, we propose the construction of the zero inflated beta prime distribution (BPIZ) from the reparametrization presented in BOURGUIGNON et al. (2018). We also determined maximum likelihood estimators and confidence intervals for the model BPIZ. We evaluate numerically estimators and confidence intervals where we check its asymptotic eficiency. Finally, we performed an application to check also its applicability in relation to real data.