Next: #das98j###
Up: Inference of finite-state machines
Previous: #manolios94j###
  Contents
  Index
(http://www.dlsi.ua.es/~mlf/nnafmc/papers/tino95learning.pdf) use a first-order DTRNN which is
basically an augmented version of
the recurrent error propagation network of Robinson and Fallside (1991), first-order DTRNN (see
section 3.2.1), with an extra
layer to compute the
output, to learn the transduction tasks performed by Mealy
machines (see section 2.3.1). The
network is trained using an online algorithm
similar to RTRL (see
section 3.4.1); weights are updated after each symbol
presentation (online learning). In addition to being one of the few
papers dealing with transduction instead of recognition tasks, it
introduces a new FSM extraction method based on Kohonen's
self-organizing
maps (see section 5.4.3; see also
Haykin (1998, 408)).
Debian User
2002-01-21