基于粗糙集解决BP网络模式识别中的噪声

为了解决BP网络模式识别中的噪声问题,提出了利用粗糙集的上、下近似和边界线集理论对噪声样本进行处理的方法。该方法对处于下近似集内的含噪声属性值,将噪声消除后转换为理想状态下的属性值;对处于边界域内的含噪声属性值保持不变。当属性值处于边界域内属性的个数与全部属性数的比值达到某个确定的值时,就认为该样本受到噪声干扰过大,对其拒绝识别。通过实验对比表明,该方法能有效地降低BP网络模式识别的误识率。     

F-二重粗积分与边界厚度

F--二重粗积分是由函数双向S-粗集生成的积分对,它是F -二重粗积分和F^--二重粗积分的推广。给出F-二重粗积分边界厚度的概念,讨论了边界厚度的动态变化规律,得到边界厚度最值定理。利用边界厚度可以对F-二重粗积分进行辨识,给出F-二重粗积分辨识定理和两个辨识原则。     

模糊粗糙集分解定理的讨论(Ⅰ)

分解定理是模糊数学的最基本定理之一,给出了模糊粗糙集的四个分解定理,并指出因偏序关系普通模糊集的部分分解定理并不能推广到模糊粗糙集。分解定理建立了模糊粗糙集与经典集间的联系,通过它可以把模糊粗糙集的某些性质的研究转化为相关的经典集的有关性质的研究,从而可以利用人们千百年来的研究成果和实践经验。     

The Hiding Characteristic of F-interior Hiding Image and Its Applications

, By using function one direction S-rough sets, the concept of F-interior hiding image is presented; the theorem of F-interior hiding and the recognition criteria of interior hiding are proposed;and the applications of F-interior hiding image are given. F-interior hiding image is a new application area of function S-rough sets, and function S-rongh sets is a new theory and new tools for iconology research.     

一类变形的McMullen集的维数及其应用

研究了平面上一类变形的Mc Mullen集R=∑∞k=1a00b-kxkyk,(xk,yk)R,其中整数a,b满足|a|≥|b|1或者|b|≥|a|1,有限整数点集R{(i,j),i=0,1,…,n-1,j=0,1,…,m-1},得到了这类自仿射集的Hausdorff维数和Box维数的计算公式.并且作为其应用给出了自仿射集R=∑∞k=1a bb a-kxkyk,(xk,yk)R相应的Hausdorff维数和Box维数,其中整数a,b满足|a-b|≥|a+b|1或者|a+b|≥|a-b|1有限整数点集R{(i+j,-i+j),i=0,1,…,|a-b|-1,j=0,1,…,|a+b|-1}.     

Machine intelligence,rough sets and rough-fuzzy granular computing:uncertainty handling in bio-informatics and Web intelligence

Machine intelligence,is out of the system by the artificial intelligence shown.It is usually achieved by the average computer intelligence.Rough sets and Information Granules in uncertainty management and soft computing and granular computing is widely used in many fields,such as in protein sequence analysis and biobasis determination,TSM and Web service classification Etc.     

三分康托集与其两个平移交的维数(英文)

C为三分康托集,考虑何时交集C∩(C+t)∩(C+s)非空,计算出当交集非空时(t,s)的Hausdorff维数.证明了:对于平面上几乎处处的(t,s),dim_HC∩(C+t)∩(C+s)=0.利用Moran集的相关结论得到当交集非空时dim_H C∩(C+t)∩(C+s)的表达式.     

Attribute Reduction in Decision Systems Based on Relation Matrix

<正>This paper proposes,from the viewpoint of relation matrix,a new algorithm of attribute reduction for decision systems.Two new and relative reasonable indices are first defined to measure significance of the attributes in decision systems and then a heuristic algorithm of attribute reduction is formulated.Moreover,the time complexity of the algorithm is analyzed and it is proved to be complete.Some numerical experiments are also conducted to access the performance of the presented algorithm and the results demonstrate that it is not only effective but also efficient.     

Pessimistic Rough Decision

<正>With granular computing point of view,this paper first presents a novel rough set model with a multigranulation view,called pessimistic rough decision,where set approximations are defined through using consistent granules among multiple granular spaces on the universe.Then,we investigate several important properties of the pessimistic rough decision model.With introduction of the rough set model,we have developed two types of multigranulation rough sets(MGRS):optimistic rough decision and pessimistic rough decision. These multigranulation rough set models provide a kind of effective approach for problem solving in the context of multi granulations.     

基于新息正交性抗野值的自适应的α-β-γ滤波跟踪器

针对传统α-β-γ滤波器抗野值能力差和受跟踪精度与收敛速度之间的矛盾以及系数固定的影响而难以跟踪强机动性目标的局限性,设计了一种基于新息正交性抗野值的自适应的α-β-γ滤波跟踪器。该滤波器通过判断α-β-γ滤波新息序列估计值的正交性质是否丧失而判别出观测值中是否存在野值,采用活化函数对新息序列进行加权修正,利用活化函数的性质,恢复新息序列原有的统计特性和正交性,从而达到抗野值的目的。同时,根据目标机动性的变化,利用跟踪误差调整模糊集系统的输出,在保证精度的前提下,得到自适应的系数值,使跟踪结果快速收敛,很好地解决了跟踪精度与收敛速度之间的矛盾。计算机仿真结果表明,在满足所要求精度的前提下,该算法可以克服传统算法的局限性,解决野值剔除问题,并有效跟踪强机动性目标。