Bayesian hierarchical models are becoming increasingly popular in analyzing complex data, particularly joint models that combine count and binary outcomes. One such model is the Joint Zero-Inflated Poisson (ZIP) mixed regression and the binomial logistic mixed regression model, which assumes a ZIP distribution for the count outcome and a logistic distribution for the binary outcome. This model also takes into account random effects to account for the correlation between observations. Another similar model is the Joint Zero-Inflated Negative Binomial (ZINB) mixed regression and the binomial logistic mixed model, which assumes a ZINB distribution for the count outcome. These models are useful for analyzing data with excess zeros and overdispersion, and can provide insights into the relationship between count and binary outcomes. Bayesian methods allow for flexible modeling of complex data structures and can provide more accurate and reliable estimates of model parameters. The main aim of the Bayesian hierarchical models is to comprehensively handle the joint analysis of count and binary data, while also accounting for covariates and addressing the issue of zero inflation. The study found that the Zero-Inflated Negative Binomial mixed model was the best fit for predicting zeros in infant mortality rates in Ethiopia. The joint model, which included factors like residence, births, mother's education, and economic status, had the best fit and lower complexity. The study concluded that the joint zero-inflated Negative Binomial mixed regression and logistic mixed regression model provided the best balance between model fit and complexity.

Mekuanint has doctoral studies from the University of Science and Technology of China. He has published more than 6 papers in reputed journals