1 paper
Daniel Paulin, Victor Elvira
Vector autoregressive moving-average (VARMA) models have long been considered impractical beyond moderate dimensions: the likelihood is non-convex, the parametrization is identifie…