1994, 5(4):1-8.
Abstract:Some important re formations and extentions that the authors made for MINIX operating system, a time-sharing、 multiuser、 multitask operating system developed by A. S. Tanenbaum in 1986 and used for teaching of operating system course, are described in details in this paper, including support for 3. 5” floppy disk, utilization of extended memory, and implementation of process swapping and spooling mechanisms.
1994, 5(4):9-18.
Abstract:Data model is the foundation of the intelligent Database Systems. The Research of intelligent Extension of Data Models is aiming at improving the performance and functions of database systems from the root of data models, so that they can handle knowledge objects, capture more semantics of the application, and have the ability of deduction. In this paper, an Object-Oriented Knowledge - Database Model is presented.The concepts and characteristics, definition and manipulation language frame and architecture of the experimental system OCEANS are discussed in detail.
1994, 5(4):19-27.
Abstract:Typed lambda-calculus is a logic system, which can be used as a programming language, such as Plotkin's PCF. Plotkin contructed a model for PCF, then he considered the matching problems between the related denotational semanitics and the operational semantics-simple match and complete match. In the present paper the authors construct another model for PCF using the concept of information system which was introduced by Scott in 1982. The authors prove that the denotational semantics are related to the model and the operational semantics are completely matched.
1994, 5(4):28-34.
Abstract:his paper introduces the MKR algorithm for Military Knowledge Acquisition system. MKR is an incremental and efficient algorithm for learning from examples. In MKR, this paper made a further improvement on IDSR algorithm, raising it's efficiency.This paper also developed an anti-disturbance mechanism in MKR, extending the field in which the algorithm can be applied.
1994, 5(4):35-39.
Abstract:xisting analyses of the Neocognitron failed to discuss the dynamic characteristic during learning because they were all confined to algebraic method. This paper intruduces differentiation to analyse the Neocognitron. For the unsupervised learning, this paper derives a defferential equation to describe us, a condition of us increasing, and an inequality that the initial value of variable weight of a, b must satisfy. For fixed representation or the supervised learning, we obtain an u. function of time. We show that, in this case, learning is a process in which us approximates a coefficience independence of q and the initial value of weight a,b.
1994, 5(4):40-48.
Abstract:his paper restricts oracle Turing machine to query its oracle in an accepting computation, and obtains result: there exist infinite number of recursive sets A, B, A',B', A", B", A, B, which are not equivalent satisfying properties: P(A,q) =P(A,q+ 1 ), P(B,q ) ≠P(B,q+1)p(a',q)=P(A' ),P(B',q)≠P(B' ),NP(A",q) = NP(A",q + 1),NP(B",q) ≠ (B",q+ 1 ),NP (A,q) =NP(A), NP(B,q)≠NP(B).
1994, 5(4):49-52.
Abstract:This paper presents an algorithm of task allocation problem using Boltzmann machines. This algorithm is simulated parallelly in Transputer multiprocessor system. It is characterized with high optimization degree and computation speed.
1994, 5(4):53-59.
Abstract:This paper discusses the splitting problem of the polynomial time bounded many one degrees. The main result is that: there exists a nonzero p-m degree a such that if a is splitted by n + 1 degrees a_{0}, a_{1}, ...,a_{n} for any natural number n≥1,then there exist at lesst n different pairs (a_{i},a_{j}) (i≠j &. i,j≤n) which are not minimal paris. This generalizes Ambos-Spies' result of which asserts that there is a nonzero p-m degree which can not be splitted by any minimal pair.
1994, 5(4):60-64.
Abstract:The Graph Search problem is proved to be NP-complete by MEGIDDO et al. An algorithm for tree is also proposed by them which computes the search number in O(n) time and the search plan in O (nlog (n) ) time. This paper developes a linear algorithm through representing a search plan by edge sequence, which computes both the search number and the search plan in O(n) time.