AUTOMATA: A Classification 2)

According to M. BUNGE: "An automaton is a discrete and sequential system"

BUNGE distinguishes deterministic automata from probabilistic ones: "If lacking in spontaneity, i.e. acting only under external compulsion, the automaton is called deterministic. If inputs and internal states determine only output probabilities, the automaton is called probabilistic.

Deterministic automata are of interest mainly in technology, where highly reliable systems are desirable. Probabilistic automata are of greater scientific and philosophical interest, for natural systems, such as brains and communities, are endowed with some spontaneity and seem to have strong stochastic components" (1979, p.263)

Some years after BUNGE wrote these lines, the development of the theory of deterministic chaos blurs somewhat the above distinction. But it also explains how an automaton, or system may function, being partly deterministic and partly stochastic.

G. KLIR proposed a different classification of automata, in accordance to their possible uses:

"1. Decoder-combinatorial automata which have the same input and output variety and in which the mapping between stimuli and responses is mutually unique.

"2. Prospective automata – combinatorial automata Prospective whose output variety is smaller than the variety at the input, and in which a single response is uniquely assigned to each stimulus, whereas the converse assignment is not unique.

"3. Retrospective automata – sequential or complex random automata, the output variety of which is larger than their input variety and where a single stimulus is assigned to each response.

"4. Function generators – sequential or complex automata which, as a result of at least one stimulus, produce in their output a response that is either a determinate or a random function of time.

"5. Self-organizing automata – sequential or complex automata, in which the variety of the responses produced by a given stimulus can change (as a rule, diminish) under certain circumstances." (1965, p.37)

This last class is very important as it corresponds to all natural and artificial autopoietic systems.

2) methodology or model

Tag cloud geneated by TagCrowd



comments powered by Disqus
Concepts extracted by AlchemyAPI AlchemyAPI