算法

The main works include five parts:(1) A modified K-means algorithm for optimizing the RBPNNs' structures was proposed;(2) The RBPNNs' structure optimization method based on the orthogonal least square algorithm was proposed that can greatly reduce the computation expense for the structure optimization;(3) Using the GA to perform the whole structure optimization of the RBPNNs was proposed in this thesis, which include simultaneous optimization of the hidden centers and the corresponding controlling parameters of the kernel functions.

主要工作体现在如下三个方面：(1)提出一种改进的 K-Means 算法；(2)提出了基于递推正交最小二乘算法的结构优化算法来训练径向基概率神经网络，从而大大减小了用于结构优化的计算开销；(3)提出使用遗传算法来实现径向基概率神经网络的全结构优化，即隐中心矢量和核函数控制参数同时优化，通过新设计的编码方式、新构造的适应度函数，充分发挥了 GA 的全局搜索性能，使得所优化的径向基概率神经网络的结构趋于最简。

By iteratively combining these two procedures we achieve a way of training the configuration of RBFNN. In addition, the algorithm is very robust with respect to the noise level. Simulation experiments show that the proposed algorithms are sound. In chapter 3, which is the extension of chapter 2, the improvements of the architectures of RBFNNs and the solutions of their relating problems are discussed, Firstly, an new distance metric is advanced and a forward orthogonal least square selection procedure is applied to learning the parameters of classification function and selecting the important input nodes.

首先分析了目前存在的同时确定RBFNN结构和参数的方法的缺陷，在此基础上提出了用基于拉马克的进化学说的进化编程算法来改进算法，以克服某些缺陷；然后针对受到严重噪声污染的系统，如何提高RBFNN的泛化能力的问题，利用基于AIC的适应度函数的改进遗传算法学习结构和参数；最后介绍基于MDL原理的方法，将优化网络的结构和参数分为两个阶段：训练和进化，先自适应地改变RBFNN基函数的中心和宽度，同时训练输出线性权值，再用基于MDL原理的适应度函数的标准GA来优化隐层节点，通过交替使用这两过程达到训练RBFNN的结构和参数的目的，该算法具有较强的鲁棒性。

The algorithm can be served as the inverse operation of raster to 2DRE quadtree, but it is not direct inversion, instead, it takes other simpler measur i.e. intersection of 2DRE quadtree and code transforming method in Raster to Quadtree conversion, to make the algorithm more effective. It can also avoid possible trouble of memory overflow caused by the operation on bigger images in smaller memory.

这一算法可看作栅格—2DRE四叉树变换算法的逆变换，但不是它的"反演"，而是采用了较为简洁的求交集运算以及栅格—2DRE四叉树变换算法中的编码转换方法，使得这一算法的实现更加有效，且避免了在图像较大而内存较小的情况下可能发生的"溢出"等矛盾。

We study the properties of $BR_0$-algebra and the total complication triple I method on complete $BR_0$-algebra, and we apply the results to $R_0$-Unite interval $\overline{W}$. Not only we have simplified the proof of the results of $R_0$-type triple I method on $R_0$-Unite interval $\overline{W}$, but also we make the proof to combine with the formal deductive system for fuzzy propositional calculus. This work also explains that the $R_0$-type triple I method is a matching fuzzy inference with $B{\cal L}^*$ system.

研究了基础$BR_0$-代数的性质和基于完备基础$BR_0$-代数的全蕴涵三I算法，对一般蕴涵算子给出了三I算法解存在的一个充分条件，并将结果应用于$R_0$-单位区间$\overline{W}$，不但极大的简化了$R_0$-单位区间$\overline{W}$的$R_0$-型$\alpha$-三I算法结果的证明，而且使其证明过程与相应的模糊命题演算系统结合起来，说明了$R_0$-型三I算法是与$B{\cal L}^*$系统相匹配的模糊推理方法。

The problem of huge numbers of datas are disposed in the process of disposed three-dimensional laser scanning datas, so on the base of carefully analszing the serial algorithm of plane segmentation and multiple datasets registration in the process of disposing three-dimensional laser scan datas, according to the method of parallel programming basing on MPI, the paper used the right means of datas distributing to design and realize the parrel algorithm of plane segmentation and multiple datasets registration.

本文针对三维激光扫描数据处理过程中所处理的数据量大这一问题，在认真分析了三维激光扫描数据处理过程中平面分割和多视点配准串行算法的基础上，根据基于MPI并行程序设计方法，采用合适的数据分配方案，设计了平面分割和多视点配准的并行算法，并给出了效果图，以及并行算法加速比和效率的实际数据，最后对并行算法效果进行了分析。

The study establishes the compression of models and algorithms for vector graphics data on different basis, including on the basis of vector graphic factors of B spline curves in predecessors research; on the basis of rectangle graphic factors in main axis mode; on the basis of the application of Hough transform to polygonal line, ellipse and circle, so the study improves the models and algorithms of compression for vector graphic data.The compression for vector graphic data mainly studies the data of storing graphic factors.

本文研究的内容是：根据对以往各种曲线矢量数据压缩模型与算法的研究，建立了基于B样条的曲线矢量图形要素的数据压缩模型和算法；在前人研究的基础上，建立基于主轴模式的矩形图形要素的数据压缩模型和算法；本文作者将Hough变换应用到折线、圆及椭圆图形要素的矢量数据压缩上，在此基础上改进了基于Hough变换的折线、圆及椭圆图形要素的数据压缩模型和算法。

The emphasis of this paper is as follows: Firstly, an improved AR model and Bayesian posterior probability based clustering method is proposed and the principle and way of how to use this method for clustering time course gene expression data are elaborated.

本论文重点是：建立了一种改进的基于自回归模型和贝叶斯后验概率的动态聚类分析算法，阐述了应用该算法进行时序基因表达数据聚类分析的原理和方法；建立了一种基于自回归模型的模糊动态聚类分析算法，阐述了应用该算法进行时序基因表达数据聚类分析的原理和方法。

Based on the traditional kinds of high-level synthesis of power-driven layout algorithm, this paper presents a new process of force-driven scheduling algorithm, in which the algorithm introduce a new data structure as an expression form of list scheduling, and improves the algorithm process.

在高层次综合中基于传统的力驱动布局算法提出一种新流程的力驱动调度算法，在该算法引进一种新的数据结构作为列表调度的表达形式，并改进了算法流程。

If we know the exact value of ψ_m, we will have, for integers ψ_m, a deterministic primality testing algorithm which is not only easier to implement but also faster than either the Jacobi Sum test, the elliptic curve test, or the AKS algorithm.

如果知道ψ_m的准确值，那么对小于ψ_m的整数N，我们就有一个确定性素性测定算法，它不仅容易实现而且比Jacobi-Sum算法、椭圆曲线素性证明算法和AKS算法速度都要快。

Finally, base on research of genetic algorithm and rough sets, analysis of the existing primary algorithm of attribute reduction based on the traditonal genetic algorithm, an improved algorithm of rough set attribute reduction base on dependability and genetic algorithm is presented in this paper.

最后，对遗传算法和粗糙集理论研究的基础上，通过分析现有的约简算法，提出一种基于依赖度的属性约简改进算法，改进后的算法可以求取属性的一个约简。

But we don't care about Battlegrounds.

但我们并不在乎沙场中的显露。

Ah! don't mention it, the butcher's shop is a horror.

啊！不用提了。提到肉，真是糟透了。