<ul><li>Variational inference (VI) is a popular method to estimate statistical and econometric models.</li><li>This paper proposes using vector copulas to capture dependence between the blocks parsimoniously.</li><li>Tailored multivariate marginals are constructed using learnable cyclically monotone transformations.</li><li>The approach demonstrates efficacy and versatility in producing more accurate posterior approximations than benchmark VI methods.</li></ul>

Vector Copula Variational Inference and Dependent Block Posterior Approximations

Discover more