A new class of ε-optimal learning automata
โ Scribed by Papadimitriou, G.I.; Sklira, M.; Pomportsis, A.S.
- Book ID
- 117938081
- Publisher
- IEEE
- Year
- 2004
- Tongue
- English
- Weight
- 343 KB
- Volume
- 34
- Category
- Article
- ISSN
- 1083-4419
No coin nor oath required. For personal study only.
๐ SIMILAR VOLUMES
A new class of automata networks is defined. Their evolution rules are determined by a probability measure p on the set of all integers 7/and an indicator function IA on the interval [0, 1]. It is shown that any cellular automaton rule can be represented b3 a (nonunique) rule formulated in terms of
A class of machines modelled as deterministic Moore automata is proposed. Particular attention is given to investigating their capability of exhibiting learning behaviour assessed by an associated index of performance. This is done by assuming that the automata are operating in a random environment