Showing papers on "ω-automaton published in 1982"
TL;DR: In theoretical physics the concept of “field” is often applied to various phenomena in the space, normally represented by differential equations, but in contrast to that, the theory of automata operates with discrete states, in which the digitalization of procedures is an important aspect.
Abstract: In theoretical physics the concept of “field” is often applied to various phenomena in the space, normally represented by differential equations. In contrast to that, the theory of automata operates with discrete states. In this, the digitalization of procedures is an important aspect. Cellular automata allow the construction of “moving state structures” representing digital particles, which may be compared with the behavior of physical particles. The theory of automata further presupposes certain attitudes towards determination and causality. In close connection is the problem of the reversibility of time direction.
59 citations
01 May 1982
TL;DR: This paper describes some experiments in applying hill-climbing to modify finite automata to accept a desired regular language, and shows that many problems can be solved by this simple method.
Abstract: : The problem addressed in this paper is heuristically-guided learning of finite automata from examples. Given positive sample strings and negative sample strings, a finite automaton is generated and incrementally refined to accept all positive samples but do no negative samples. This paper describes some experiments in applying hill-climbing to modify finite automata to accept a desired regular language. We show that many problems can be solved by this simple method. We then describe the method how to 're-construct' a finite automaton if the positive and/or negative samples are slightly altered, without starting from the beginning. Finally, we have an actual system. RR: Regular set Recognizer, that learns to recognize a regular set from the samples that are given by a human teacher one by one.
45 citations
27 Sep 1982
TL;DR: An algorithm to construct distributed systems from cycle-free finite automata and a partition of the input-alphabet is introduced.
Abstract: The purpose of this paper is to introduce an algorithm to construct distributed systems from cycle-free finite automata and a partition of the input-alphabet.
4 citations
TL;DR: Several interpretations of Theorem 2 are discussed which offer some insight into some mathematical limits of machine intelligence.
Abstract: Computers and brains are modeled by finite and probabilistic automata, respectively. Probabilistic automata are known to be strictly more powerful than finite automata. The observation that the environment affects behavior of both computer and brain is made. Automata are then modeled in an environment. Theorem 1 shows that useful environmental models are those which are infinite sets. A probabilistic structure is placed on the environment set. Theorem 2 compares the behavior of finite (deterministic) and probabilistic automata in random environments. Several interpretations of Theorem 2 are discussed which offer some insight into some mathematical limits of machine intelligence.
2 citations
Book•
01 Dec 1982
TL;DR: The automata superponable with respect to pairs of operations and Invariant relations of automata are presented.
Abstract: 1 Automata - definitions and notations- 2 Linear automata- 3 Automata superponable with respect to pairs of operations- 4 Automata superponable with respect to pairs of automata- 5 Invariant relations of automata
2 citations
TL;DR: Simulation results indicate that by a proper choice of the updating functions, the automata converge to optimum parameter values, and the game approach appears to be one way of reducing the high dimensionality of decision space.
2 citations
12 Jul 1982
TL;DR: The class of languages accepted by the one-way nondeterministic simple k-head finite automata ℒ(NSPk-HFA) is not closed under concatenation for any k≥2, and the class of Languages recognized by one- way k- head deterministic sensing finite state automataℒ (1DSeFA(k)) are not close under Concatenation, Kleene star and reversal.
Abstract: The following results are shown :
(1)
The class of languages accepted by the one-way nondeterministic simple k-head finite automata ℒ(NSPk-HFA) is not closed under concatenation for any k≥2.
(2)
The class k U ℒ(NSPk-HFA) is closed under concatenation.
(3)
The class of languages recognized by one-way k-head deterministic finite state automata ℒ(1DFA(k)) and the class of languages recognized by one-way k-head deterministic sensing finite state automata ℒ(1DSeFA(k)) are not closed under concatenation, Kleene star and reversal.
1 citations
TL;DR: The study of the characteristic properties of abstract finite computer automata and their growing sequences leading to infiniteComputer automata will clarify the deep and crucial differences between traditional computation theory based on an infi'n~te concept Ce and a computation theory for computers which are strictly finite devices.
Abstract: The study of the characteristic properties of abstract finite computer automata and their growing sequences leading to infinite computer automata will clarify the deep and crucial differences between traditional computation theory based on an infi'n~te concept Ce.g., a Turing machine) and a computation theory for computers which are, and always will be, strictly finite devices. THe study of concrete arithmetic automata and programming languages accepted by ~m ~ will provide insight rinto the essential properties of finite computer arithmetics and their axiomatic-systems with respect to proving the correctness of programs an~ will lead to a knowledge which could ~nfluence the design of a new computer arithmetics and new architectures. The programming language accepted ~y a finite concrete computer automaton is an advanced defini'tion as it reflects the most important property of programs, i.e., teat-they are algorithms which should be executed. A domain and function are associated wit~ each program.-Further, the execution sequence and computat¢on are clearly differentiated. A computer automaton differs from a classical automat6n by viewing input symbols as instructions, and states as states of memory. Thus, each computer automaton controlled by a sequence of in$~u ctions may define a function from states to states. 22 I. Abstract Finite Computer Automaton A__n abstract finite computer automaton., (process) on A ~_Z Alg with s~. The programming (algorithmic) language, ALA, accepted by A is the set of all programs Alg such that there exists at least one st~ and CActiv(A,Alg,st~). The function, FA,A~ (st&)=StK, for each st~eDomaigAlg is called evaluatable on A b~ Alg. Both Alg and Alg* from A1A are function equivalent if FA,A~ =FA,i~. If ProcessA, i~ is the set of all completed execution sequences, Alg and Alg* are process equivalent if ProcessA,~ =Process Aj~t~e. 23 function and process equivalencies are decidable, the former is much less efficient than the latter. A finite abstract computer automaton can be represented in plane as its transition graph, TGA=, where • is a labelling of edges, (st,st*)&T~StxSt, by sets of instructions defined as follows: r=(st,st*)={in~ In;tra___D_~(in,st)=st*]. edges, (in,in~)~C~PxP, by sets of states defined as follows: ~(in,in*)={st~St;co~(in,st)=in*}. Graph theoretical terminology may then be used. Each completed computation is a finite path in a transition graph while each completed execution sequence (process) is a finite path in a control graph starting with inK, etc. If contr~ in Alg=. is independent of states, Alg is called sequen%i@! and the multiset, P, may be ordered in …