EM Estimation for Zero- and k-Inflated Poisson Regression Model

Count data with excessive zeros are ubiquitous in healthcare, medical, and scientific studies. There numerous articles that show how to fit Poisson other models which account for the zeros. However, many situations, besides zero, frequency of another count k tends be higher data. The zero- k-inflated distribution model (ZkIP) is appropriate such situations ZkIP essentially a mixture degenerate distributions at points zero k. In this article, we study fundamental properties distribution. Using stochastic representation, provide details obtaining parameter estimates regression using Expectation–Maximization (EM) algorithm given We derive standard errors EM by computing complete, missing, observed information matrices. present analysis two real-life methods outlined paper.