A general Bayesian model for heteroskedastic data with fully conjugate full-conditional distributions

Models for heteroskedastic data are relevant in a variety of applications ranging from financial time series to environmental statistics. However, the topic modelling variance function conditionally has not seen as much attention mean. Volatility models have been used specific applications, but these can be difficult fit Bayesian setting due posterior distributions that challenging sample efficiently. In this work, we introduce general model data. This approach conditional any desired covariates or random effects. We rely on multivariate log-Gamma distribution theory construct priors yield fully conjugate full-conditional Gibbs sampling. Furthermore, extend deep learning provide highly accurate estimates dependent also an extension heavy-tailed illustrate our methodology via three applications.