Structure and parameter learning via entropy minimization, with applications to mixture and hidden Markov models.

Matthew Brand

Browse the full ICASSP paper archive.