Closed-Form Predictive Density Estimation for Bivariate Gamma Distribution With Application in Hydrological Flood Data Estimación de densidad predictiva en forma cerrada para la distribución Gamma bivariada con aplicación en datos hidrológicos de inundaciones

Finding closed-form solutions in Bayesian data analysis can be critical and time-saving, as it eliminates the need for computationally expensive techniques like MCMC methods. This paper explores Bayesian analysis with closed-form solutions of the bivariate gamma distribution. We present predictive density estimations under the Kullback-Leibler divergence, utilizing three well-known (non-) informative prior distributions, all analyzable in closed form. We compare these methods through simulation studies and a real-world example, applying them to hydrological flood data.