Confidence Interval Estimation for the Common Mean of Several Zero-Inflated Gamma Distributions

In this study, we propose estimates for the confidence interval common mean of several zero-inflated gamma (ZIG) distributions based on fiducial generalized (GCI) and Bayesian highest posterior density (HPD) methods Jeffreys rule or uniform prior. Their performances in terms their coverage probabilities expected lengths are compared via a Monte Carlo simulation study. For almost all scenarios considered, results show that GCI performed better than HPD methods. Daily rainfall data from Chiang Mai Province, Thailand contains zero entries follows ZIG distribution is used to test efficacies real-world situations.