一种基于神经网络的分形几何图的产生与编码

本文提出用概率逻辑神经网产生一类自（互）相似（分形）图的方法，指出这类图能用一组压缩编码表示，给出快速寻找该编码的算法，既识别该分形几何图的算法，证明该编码是最优的，即码的长度最短，这些成果有希望在图象压缩和模式识别中得到应用。     

Undecidability of the trace coding problem and some decidable cases

We introduce and investigate the notion of weak morphisms of trace monoids with the aim of dealing with the problem of deciding the existence of codings between trace monoids. We prove that this problem is not recursively enumerable, which answers the question raised by Ochma ski in 1988. On the other hand, we show its decidability when restricted to instances with domain monoids defined by acyclic dependence graphs. We also partially answer the question of Diekert from 1990 about the number of free monoids needed for encoding a given trace monoid into their direct product.     

Nonterminals,homomorphisms and codings in different variations of OL-systems

Summary Continuing the work begun in Part I of this paper, we consider now variations of nondeterministic OL-systems. The present Part II of the paper contains a systematic classification of the effect of nonterminals, codings, weak codings, nonerasing homomorphisms and homomorphisms for all basic variations of non-deterministic OL-languages, including table languages.Part I of this paper appeared in Acta Informatica 4/1 (1974).     

Visual coding by optimal graph-coloring

A formal study of visual codings in user interface design is presented. Visual codings for maximum distinction of different objects in displayed images are formulated as a discrete optimization problem of maximum-distance graph-coloring. The formulation is a generalization of the classical coloring problem in graph theory. Having pointed out that maximum-distance graph-coloring is NP-complete, we develop new, fast approximation algorithms for optimal visual codings. The proposed algorthms run inO (M N) time, whereM is the number of visual codes used andN is the number of objects to be encoded. Besides being efficient, the algorithms are simple and easy to implement. Our experiments showed that geographic maps automatically colored by the new algorithms were preferred to those colored by the previous graph-theoretical approach and they are competitive, if not better, in terms of the visual distinction of different regions than those drawn by hand.Supported by grans from Natural Science and Enginneering Research Council of Canada     

Nonterminals,homomorphisms and codings in different variations of OL-systems

Summary The use of nonterminals versus the use of homomorphisms of different kinds in the basic types of deterministic OL-systems is studied. A rather surprising result is that in some cases the use of nonterminals produces a comparatively low generative capacity, whereas in some other cases the use of nonterminals gives a very high generative capacity. General results are obtained concerning the use of erasing productions versus the use of erasing homomorphisms. The paper contains a systematic classification of the effect of nonterminals, codings, weak codings, nonerasing homomorphisms and homomorphisms for all basic types of deterministic OL-languages, including table languages.     

Homomorphic images of sentential form languages defined by semi-Thue systems

We investigate various language families which can be obtained from sentential form languages of semi-Thue systems by applying codings, weak codings, non-erasing homomorphisms, and arbitrary homomorphisms. We also distinguish between monotone, context-independent, and context-dependent semi-Thue systems with at most one or finitely many starting words. Finally, we study the effect of erasing productions u → λ.We obtain complete diagrams which show inclusions and incomparabilities of all the language families so defined.     

Online error correcting output codes

This article proposes a general extension of the error correcting output codes framework to the online learning scenario. As a result, the final classifier handles the addition of new classes independently of the base classifier used. In particular, this extension supports the use of both online example incremental and batch classifiers as base learners. The extension of the traditional problem independent codings one-versus-all and one-versus-one is introduced. Furthermore, two new codings are proposed, unbalanced online ECOC and a problem dependent online ECOC. This last online coding technique takes advantage of the problem data for minimizing the number of dichotomizers used in the ECOC framework while preserving a high accuracy. These techniques are validated on an online setting of 11 data sets from UCI database and applied to two real machine vision applications: traffic sign recognition and face recognition. As a result, the online ECOC techniques proposed provide a feasible and robust way for handling new classes using any base classifier.     

MPEG-2 Coded Video Traces Transmitted Over a Satellite Link: Scalable and Non-Scalable Solutions in Rain Fading Conditions

The literature is poor in the analyses of the effects produced by corrupted bits in compressed video bitstreams. This paper presents the results of a transmission experiment of MPEG-2 coded video data over a satellite link affected by noise, in order to investigate under which conditions this type of transmission is economically feasible. The signal-to-noise ratio scalability feature of the MPEG-2 encoder was used to produce different bitstreams of the same movie sequence. The scope of the study was to verify which are the best combinations of video and channel codings in the presence of attenuation on the satellite link, in order to optimize the bandwidth utilisation for a requested image quality. The results obtained give indications about the data channel codings to be used to counter the rain fade on the transmission link, which is a non negligible problem especially when satellite transmissions are in the Ka band. Moreover, the results highlight the flexibility of the scalable video coding in the examined scenario.     

Semantics of trees

We introduce F-structures and investigate interpretations of trees in F-structures. Thus trees are viewed as schemes of computations in F-structures. Every tree is shown to be strongly equivalent to some chain-free tree. This enables us to use the grammatical codings of [2] to define formulas for arbitrary interpreted trees. We also introduce a syntactic way to compute the values of interpreted trees.     

具有可自适应性的问卷设计器的设计

文章首先分析了目前问卷调查系统所存在的问题,提出了改进的意见。然后从便于问卷的界面实现和结果保存关键问题出发,探讨问卷组成成分、答题顺序控制和答题结果编码等环节的解决方案,为问卷的存储、答卷界面动态生成和答卷结果实时保存的实现提供技术支持。根据设计器的技术要求,采用动态界面生成技术和编码技术,对问卷的表现形式、问卷对象的属性和问卷对象的方法进行详尽的描述,设计了问卷设计器的数据结构和处理流程,完成了可自适应性的问卷设计器的设计。     

DCT算法及其与小波编码在图像处理中的比较

由于数字媒体的不断普及,DCT编码在JPEG,MPEG,H.26x等编码标准中有着越来越广泛的应用。但随着Shapiro关于小波的嵌入式零树小波编码以及Said等人提出的集合分裂等级树编码的成功应用,图像及视频信号的压缩编码迎来了一个新的阶段,这显然对传统DCT编码提出了挑战。然而,Xiong等人利用嵌入式DCT块变换之间直流系数(DC)的高度相关性也能获得较高的编码性能,因为对DCT后的系数进行重组,同样具有小波多分辨率图像的特性。文中主要提出DCT与DWT的比较,以及基于DCT与小波变换在图像中的性能比较。     

多协议标记交换中标记分发协议的研究

多协议标记交换（ＭＰＬＳ）将面向连接的机制加入到面向非连接ＩＰ协议中,不仅能大在怕转发速度和网络的吞叶量,还十分有利于ＱｏＳ的实现,是一种非常有前景的网络技术,而标记分发协议（ＬＤＰ）是ＭＰＬＳ的核心部分。本文对ＬＤＰ的基本运作、ＬＤＰ ＰＤＵ格式及其相应参数的ＴＬＶ编码、ＬＤＰ中的消息等进行了介绍和讨论。     

Operations on logic functions and variables through sign Haar spectra

A non-linear transform, called “sign Haar transform” has recently been introduced. The transform is unique and converts binary/ternary vectors into ternary spectral domain. Recursive definitions for the calculation of sign Haar transform are developed. Essential properties of logic functions and variables in the spectral domain of a quantized transform based on Haar functions are presented. Sign Haar transform has the smallest computational cost of all the quantized transforms. The properties of logic functions are listed for two different codings of incompletely specified functions.     

Yang—Feng网络的时间计算算法研究

在表达并行过程的Yang-Feng网络的基础上，采用网络优化理论，定义和构造了一种过程时间网络，以表示和计算过程时间；并对两种不同的网络节点编号，分别研究和提出了过程时间计算方法，其适应性较广，可以很好地应用于并行设计和并行工程的有关研究领域。     

Palindromic complexity of codings of rotations

We study the palindromic complexity of infinite words obtained by coding rotations on partitions of the unit circle by inspecting the return words. The main result is that every coding of rotations on two intervals is full, that is, it realizes the maximal palindromic complexity. As a byproduct, a slight improvement about return words in codings of rotations is obtained: every factor of a coding of rotations on two intervals has at most 4 complete return words, where the bound is realized only for a finite number of factors. We also provide a combinatorial proof for the special case of complementary-symmetric Rote sequences by considering both palindromes and antipalindromes occurring in it.     

A decimal-coded evolutionary algorithm for constrained optimization

A proposed integer decimal coding is described and compared with corresponding binary codings. An evolutionary program for structural design optimization is developed based on a decimal coding of the design variables. It is shown that decimal coding, processed using ranking, elitism and mutation, provides a satisfactory computational basis and offers certain advantages over binary versions. Two types of mutation are used, one random and the other `intelligent' or beneficial. The program is applied to constrained numerical minimization problems and to constrained structural minimum weight evaluations.     

The Rotation χ-Lattice of Ternary Trees

This paper generalizes to k-ary trees the well-known rotation transformation on binary trees. For brevity, only the ternary case is developped. The rotation on ternary trees is characterized using some codings of trees. Although the corresponding poset is not a lattice, we show that it is a χ-lattice in the sense of Leutola–Nieminen. Efficient algorithms are exhibited to compute meets and joins choosen in a particular way. Received May 22, 2000; revised December 11, 2000     

EOL languages are not codings of FPOL languages

One of the useful results concerning EOL languages states that a language is an EOL language if and only if it is a cording of OL language. In this paper we retine this result by demonstrating that there exist EOL languages that are not codings of languages that are generated by propagating OL systems with finite axiom sets. This solves Problem 10 from the L Systems Problem Book '75.     

On the structure of certain intersection graphs

Recently Gavril introduced a new class of intersection graphs called interval-filament graphs. These include co-comparability graphs and polygon-circle graphs (the intersection graphs of polygons inscribed in a circle), which include circular-arc graphs (the intersection graphs of arcs of a circle), circle graphs (the intersection graphs of chords of a circle), chordal graphs, and outerplanar graphs. We give a structural property of polygon-circle graphs. We prove a bound on the clique-covering number for interval-filament graphs in terms of the size of a largest independent set of nodes in the graph. We prove that complements of interval-filament graphs are 4-divisible in the sense of Hoàng and McDiarmid. Similar results are obtained for complements of other intersection graphs introduced by Gavril.     

Sticker systems with complex structures

 In this paper, we propose a variant of sticker systems which uses molecules with complex structures. Since the original sticker systems (Paun et al. (1998) [2, 8]) working on double strands of DNA have been studied as a formal model for self-assembly in DNA computing, we extend the sticker systems to working on more complex (higher-order) structures of DNA molecules. The advantage of sticker systems with complex structures is that augmented with weak codings we can obtain the characterization of recursively enumerable languages by using only sticking (hybridization) operations for complex molecules, while the usual sticker systems require more complicated operations such as the simultaneous use of couples of dominoes or coherent computations besides morphisms.