Automatic selection of acoustic and non-linear dynamic features in voice signals for hypernasality detection | Publicación