Next: Turing computability with DTRNN
Up: Featured papers
Previous: #omlin96j3###
  Contents
  Index
As has been discussed earlier in
this chapter, certain DTRNN architectures may not be capable of
representing all possible FSM. Recurrent cascade correlation (RCC) (see section 3.4.3) is
both the name of a learning algorithm which constructs a DTRNN during
learning and the name of the class of architectures generated by this
algorithm. A number of papers
have addressed the representational capabilities of RCC networks
((Giles et al., 1995), (Kremer, 1996a),; (Kremer, 1996b)) by
defining a class of FSM that cannot be represented in RCC nets; the
class grows in generality as we go from one paper to another. A
general formulation is presented by Kremer (1996b) (http://www.dlsi.ua.es/~mlf/nnafmc/papers/kremer96finite.pdf), but the class of FSM that can be represented in this
architecture still remains to be defined. According to
Kremer (1996b), which uses a reductio ad
absurdum proof, RCC nets cannot represent FSM which, as
a response to input strings whose symbols repeat with a periodicity
, output strings have a periodicity such that
(if is even) or
(if is
odd). One such automaton is the odd parity automaton (see
figure 2.3), whose output has a period if its input is a
constant ``11111...'' (period ).
Next: Turing computability with DTRNN
Up: Featured papers
Previous: #omlin96j3###
  Contents
  Index
Debian User
2002-01-21