中国科学院数学与系统科学研究院期刊网

1996年, 第9卷, 第3期 刊出日期:1996-07-15
  

  • 全选
    |
    论文
  • YUAN Jinjiang;ZHANG Heping
    Journal of Systems Science and Complexity. 1996, 9(3): 193-197.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    The computational complexity of the minimum fill-in problem of graphs is discussed in this paper. We prove that this problem is NP-complete even for bipartite graphs and square graphs. A linear time algorithm for outerplane graphs is got. The computational complexity of the minimum fill-in problem of graphs is discussed in this paper. We prove that this problem is NP-complete even for bipartite graphs and square graphs. A linear time algorithm for outerplane graphs is got.
  • WU Jinzhao
    Journal of Systems Science and Complexity. 1996, 9(3): 198-204.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    For the monoids presented by finite special Church-Rosser Thue systems,the number of each kind of Green equivalence classes is either one or infinite, and it is computable in polynomial time. As applications, the regular problem and the existence of non-trivial idempotents for such monoids are decidable in polynomial time. And a result about the descriptive power of finite special Church-Rosser Thue systems is obtained.
  • ZHOU Yong
    Journal of Systems Science and Complexity. 1996, 9(3): 205-215.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    For the nonparametric smooth PL-estimator Fn, the smooth PL-process (Fn - F) can be uniformly represented by a smooth empirical process plus a negligible remainder term when the data are subject to random left truncation. The law of the iterated logarithm (LIL) and the weak convergence of the smooth PL-process are established from this representation.
  • WANG Dongming
    Journal of Systems Science and Complexity. 1996, 9(3): 216-228.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    This is the first in a series of papers describing an elimination method for differential (abbr. d-) polynomial systems. We present two mds algorithms, one with and the other without projection, for decomposing any system of multivariate ordinary d-polynomials over an ordinary d-field of characteristic 0 into d-triangular systems. The second algorithm that computes d-triangular systems possessing certain projection property also yields a quantifier elimination procedure for algebraically closed d-fields. An example is given to illustrate our algorithms with their preliminary implementation in Maple. In the forthcoming papers of this series, we shall present algorithms for irreducible zero decomposition of ordinary d-polynomial systems and zero decomposition of partial d-polynomial systems over d-fields of arbitrary characteristic.
  • CHEN Guangya
    Journal of Systems Science and Complexity. 1996, 9(3): 229-235.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    A kind of equilibrium problems of networks with vector-valued cost functions has proposed. Equivalence between the equilibrium problem of networks and the vector variational inequality is proved.
  • HUANG Sha
    Journal of Systems Science and Complexity. 1996, 9(3): 236-241.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    In this paper, introducing the concept of“quasi- permutation”,we solve the sign problem often met with when the operation is carried out in the incommutable Clifford algebra, and using the results in Hua Luogeng and I. N. Vekua famous works, we give the existence and uniqueness theorem of solution of the two problems of Dirichlet's type for the regular functions with values in a 2^{n-1}-dimensional Clifford algebra An(R) in the hyperball and the integral representation of the solution of this problem.
  • GAO Hongzhu
    Journal of Systems Science and Complexity. 1996, 9(3): 242-251.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    Let M be a nonorientable surface of genus n. M embeds in a four-manifold N which is not necessarily homologically 1-connected. In the case of M representing zero class in H_2(N,Z_2), restrictions on normal Euler numbers of these embeddings are given.
  • SHI Jian
    Journal of Systems Science and Complexity. 1996, 9(3): 252-258.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    In the literature of resampling methods, the bootstrap and random weighting are much more studied (the latter is also called Bayesian bootstrap). When they are applied to approximate distributions of pivotal statistics, the strong consistency is often discussed.However, one could not determine which is better becauae both of their accuracies are O (n/(loglogn)1/2). In this note, a comparison between them is investigated under the Lp norm. A result obtained shows that the bootstrap is relatively better.
  • XU Dachuan;LIU Guanghui
    Journal of Systems Science and Complexity. 1996, 9(3): 259-269.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    The DFP algorithm is one of the mest famous numerical algorithms for unconstrained optimization.It is an open problem whether DFP algorithm with inexact line searches is globally convergent. In this paper, we propose a sufficient condition for the convergence of the DFP algorithm with Wolfe line search on uniformly convex objective function.
  • LIU Xiaoqi
    Journal of Systems Science and Complexity. 1996, 9(3): 270-277.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    In this paperl we show that the finite element solution of parabolic equation hasa multi-parameter asymptotic error expansion. From this expansion, a multi-parameter splitting extrapolation method and a Parallel algorithm for solving parabolic equations with higher order accuracy are developed.
  • LIANG Hua;CHEN Guijing
    Journal of Systems Science and Complexity. 1996, 9(3): 278-282.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    This paper is concerned with the asymptotic efficiency for two classes oftruncated distributions: f(x;\theta)I_{(\theta\leq x\leq 2\theta)}, and f(x;\theta)I_{(\theta\leq x\leq \theta+1)}. It is shown that the usual estimators of are asymptotically medially unbiased, but noneis asymptotically efficient in the sense of Takeuchi.
  • J.A.Muller
    Journal of Systems Science and Complexity. 1996, 9(3): 283-287.
    摘要 ( ) PDF全文 ( )   可视化   收藏
    Several methodological Problems of experimental systems analysis can be overcome by means of parametric and nonparametric selection procedures. Together with theselection of models with optimal complexity and application of optimal clusterizations forstructure analysis the selection of Predictions by means of analogous patterns is considered.