Paper available

[Mikel L. Forcada]

Dear connectionists:

The following paper, to appear in Neural Computation is now available
via our web server:

               PDF: http://www.dlsi.ua.es/~mlf/stanc3web.pdf
gzipped PostScript: http://www.dlsi.ua.es/~mlf/stanc3web.ps.gz


                    Stable encoding of finite-state machines 
              in discrete-time recurrent neural nets with sigmoid units

           Rafael C. Carrasco, Mikel L. Forcada, M. Ángeles Valdés-Muñoz
                 Departament de Llenguatges i Sistemes Informàtics
                              Universitat d'Alacant
                              E-03071 Alacant (Spain)

                                  Ramón P. Ñeco
                Departament de Ciències Experimentals i Tecnologia
                          Universitat "Miguel Hernández"
                                E-03202 Elx (Spain)

                                     ABSTRACT
 
In recent years, there has been a lot of interest in the use of
discrete-time recurrent neural nets (DTRNN) to learn finite-state tasks,
with interesting results regarding the induction of simple finite-state
machines from input--output strings. Parallel work has studied the
computational power of DTRNN in connection with finite-state
computation. This paper describes a simple strategy to devise stable
encodings of finite-state machines in computationally capable
discrete-time recurrent neural architectures with sigmoid units, and
gives a detailed presentation on how this strategy may be applied to
encode a general class of finite-state machines in a variety of
commonly-used first- and second-order recurrent neural networks.  Unlike
previous work that either imposed some restrictions to state values, or
used a detailed analysis based on fixed-point attractors, the present
approach applies to any positive, bounded, strictly growing, continuous
activation function, and uses simple bounding criteria based on a study
of the conditions under which a proposed encoding scheme guarantees that
the DTRNN is actually behaving as a finite-state machine.


-- 
_____________________________________________________________________
Mikel L. Forcada                    E-mail: mlf at dlsi.ua.es
Departament de Llenguatges          Phone: +34-96-590-3400 ext. 3384;
i Sistemes Informatics                also +34-96-590-3772.
UNIVERSITAT D'ALACANT               Fax:   +34-96-590-9326, -3464
E-03071 ALACANT, Spain.
______________________________________________________________________
URL: http://www.dlsi.ua.es/~mlf
----------------------------------------------------------------------