共查询到20条相似文献,搜索用时 78 毫秒
1.
Rough set theory is a new soft computing tool, and has received much attention of researchers around the world. It can deal with incomplete and uncertain information. Now, it has been applied in many areas successfully. This paper introduces the basic concepts of rough set and discusses its applications in Web mining. In particular, some applications of rough set theory to intelligent information processing are emphasized. 相似文献
2.
WANG Jian WANG Huiqiang ZHAO Guosheng 《武汉大学学报:自然科学英文版》2006,11(6):1785-1788
0 IntroductionNetwork survivability breaks li mits of traditional networksecurity concepts ,and emphasizes the abilities of net-work and information systems to achieve their main missionsin a ti mely manner , while they are suffering from attacks ,faults , or accidents[1]. Situation assessment[2]for networksurvivabilityis an assessment process for its dynamic changecondition. We can achieve assessment for network survivabili-ty according of the coming data.By assessing evolvement fornetwork su… 相似文献
3.
WANG Hongmei QIAN Jinping ZHANG Xiulan HE Xiubin 《武汉大学学报:自然科学英文版》2007,12(4):749-754
Aiming at providing theoretical basis for effective protection of biodiversity, the study presents a cascade method which combines both qualitative and quantitative methods, incorporates basic data with RS(remote sense) technology, and ranks the ecosystems according to its ability of biodiversity sustainability in Hebei Province. The results indicate that the most important areas for protection in Hebei Province are forest and meadow ecosystems in some highlands around Xiaowutai Mountain, Wuling Mountain, North Hebei, Taihang Mountain and East Hebei; grass ecosystems in part of plateau area and North Hebei; and some scattered wetlands in the plain and inshore areas. This method is suitable for undertaking large-scale investigations especially when the data are not adequate or unevenly distributed spatially. 相似文献
4.
WAN Qiuhong FANG Shengguo LI Jianguo ZHANG Liming OU Weifu XIAN Fanghai CHEN Wanli 《科学通报(英文版)》2005,50(17):1879-1886
In the present study, formalin-fixed feces, oligonucleotide fingerprinting and SRY-gene based sexing were used to construct a family net for giant pandas in the Tangjiahe Natural Reserve and to assess contemporary gene flow (migration) in this population. A total of 124 fecal samples were attributed to 37 individuals (22 females and 15 males) that were then analyzed for family relationships. Based on DNA fingerprints, the deduced family net revealed the following facts: (i) First-order relatives possessed similarities from 50% to 90%, and similarities between unrelated individuals or distant relatives were as high as 77%, indicating that the Tangjiahe pandas are characterized by high genetic similarity; (ii) 15 matings were identified and 5 ones occurred between close relatives, implying that there is potential for inbreeding to impact the pandas; (iii) four mating pairs and 5 offspring presented long distance migrations, demonstrating the intra-reserve habitat is continuous; (iv) four pairs of full sibs (also female-male dyads) dispersed short distance and all of them gave birth to highly inbred offspring, reflecting long distance migration is vital for inbreeding avoidance; (v) 17 adult individuals dispersed short or moderate distance and formed three clusters on the landscape, indicating that it is necessary to find out whether there is a negative factor impacting the pandas. 相似文献
5.
Da Hsuan Feng 《复旦学报(自然科学版)》2007,46(1):3-10,40
1 Preface
Thank you so much for giving me the honor to speak here today. There is a famous story about a bhnd person trying to figure out what an elephant looks like. When he touches the tail, he said the elephant looks like the tail. When he touches the ears, he said that the elephant looks like the ears. The reason is because the elephant is huge and unless you can examine both closely and afar, with open eyes, you may not see the whole Dicture. 相似文献
6.
ChenYong ChenGuo-jie WangQing 《武汉大学学报:自然科学英文版》2003,8(3B):935-939
By using the concept of emergy and method of emergy analysis, this paper has studied two mountain settlement (village) ecosystems in two aspects: input and output of energy in the agro-systems and the use of fuel energy. The result reveals that the settlement agro-system in the valley of Minjiang River is better in both structure and function than that at the so-called half-high mountain. The former one is higher than the latter one regarding such indices as emergy yield ratio, labor productivity of emergy and emergy sustainable index, but lower than the latter one regarding environmental loading ratio. In terms of fuel emergy, the settlement in the valley enjoys more diversified sources and is less depending on bio-energy while the settlement at the half-high mountain has only one source (i. e. fuel wood), much depending on bio-energy, may exert a greater pressure on environment. 相似文献
7.
WangYi-qian TanYan 《武汉大学学报:自然科学英文版》2003,8(3B):905-910
This paper studies current policies in compensation for land lost due to construction of the Three Gorges project (TGP), China. In the case study of Kaixian County, the value of land in the peri-urban areas is underestimated. Some losses in land are tangible and thus measurable, but some losses are intangible and thus immeasurable. The land available in the resettlement areas is scarce and rarely of equivalent quality. This study suggests that the loss of agricultural land be compensated with alternative land and/or non-land-based livelihood strategies, including part compensation in monetary terms and preferential policies. 相似文献
8.
ZHANGDe'er DemareeGaston 《科学通报(英文版)》2004,49(23):2508-2514
In the context of historical climate records of China and early meteorological measurements of Beijing discovered recently in Europe, a study is undertaken on the 1743 hottest summer of north China over the last 700 a, covering Beijing, Tianjin, and the provinces of Hebei, Shanxi and Shandong, with the highest temperature reaching 44.4℃ in July 1743 in Beijing, in excess of the maximum climate record in the 20th century. Results show that the related weather/climate features of the 1743 heat wave, e.g., flood/drought distribution and Meiyu activity and the external forcings, such as solar activity and equatorial Pacific SST condition are the same as those of the 1942 and 1999 heat events. It is noted that the 1743 burning summer event occurs in a relatively warm climate background prior to the Industrial Revolution, with a lower level of CO2 release. 相似文献
9.
Geochemical characteristics and petrogenesis of Mesozoic basalts from the North China Craton: A case study in Fuxin, Liaoning Province 总被引:13,自引:0,他引:13
ZHANGHongfu ZHENGJianping 《科学通报(英文版)》2003,48(9):924-930
Occurrence of Cretaceous basalts in Fuxin County, Liaoning Province provides us an opportunity to understand Mesozoic mantle processes beneath the northern margin of the North China Craton (NNCC). Fuxin Jianguo basalts occur as volcanic channel phases with well-developed columnar jointings and contain few spinel lherzolite and pyroxenite xenoliths. They are poor in silica and rich in alkalis, Ti and Al, belonging to alkaline basalts. In trace element compositions, Jianguo basalts are moderately enriched in LREE and LILE, but not depleted in HFSE. They have low Sr and high Nd and Pb isotopic ratios. These geochemical characteristics suggest that Jianguo basalts originated from the depleted asthenosphere, representing an undifferentiated and uncontaminated primitive magma. Presence of these basalts indicates that the lithosphere beneath the region had thickness less than 65 km at the time of basalt eruption and was mainly composed of fertile pargasite-bearing spinel Iherzolite and plagioclase pyroxenite. The voluminous basaltic-andesitic magmatism during the early Jurassic-late Cretaceous time indicates that the commencement and accomplishment of lithosphere thinning in the NNCC was much earlier than that in the southern margin, since the mafic-intermediate volcanism only occurred at the Cretaceous time in the southern margin and the basalts with an asthenosphere isotopic signature at the Tertiary. This shows that highly spatial and temporal heterogeneity existed in the Mesozoic lithosphere evolution. 相似文献
10.
FU Xudong WANG Guangqian KANG Zhicheng FEI Xiangjun 《武汉大学学报:自然科学英文版》2007,12(4):583-587
Characteristics of planar velocity distribution of viscous debris flow were analyzed using the measured data at Jiangjia Ravine, Yunnan, China. The velocity data were measured through using two radar velocimeters. The cross-sectional mean velocities were calculated and used to examine Kang et al's (2004) relationship, which was established for converting the flow velocity at river centerline measured by a radar velocimeter into the mean velocity based on the stop-watch method. The velocity coefficient, K, defined by the ratio of the mean velocity to the maximum velocity, ranges from 0.2 to 0.6. Kang et al's (2004) relationship was found being inapplicable to flows with K smaller than 0.43. This paper contributes to show the complexity of the planar velocity distribution of viscous debris flows and the applicability of Kang et al's relationship. 相似文献
11.
面对生活中数据信息量大的问题,常使用粗糙集对数据进行知识约简,消除数据中冗余的部分。但大多数研究没有考虑约简后对原有分类的影响;常使用的朴素贝叶斯算法又难以获得其先验概率。基于上述问题,本文提出了一种基于粗糙集的贝叶斯分类算法:首先利用粗糙集中决策属性和条件属性之间的依赖关系,进行属性约简,消除冗余的数据,然后通过贝叶斯算法对约简后的数据进行知识挖掘,最后通过对故障源数据的对比分析。该方法既避开了朴素贝叶斯算法对先验概率的要求,又使得数据分类和预测能力有了明显提升。 相似文献
12.
要从数据量庞大的大型数据库中发现知识,就不得不考虑学习算法的效率。将粗集理论应用到数据挖掘中,实现了从数据库中自动抽取与给定的学习任务相关的属性,能有效地发现简练、贴切的知识,并给出了这一应用的理论基础和实现算法。 相似文献
13.
一种基于粗糙集理论的数据挖掘算法的研究 总被引:6,自引:0,他引:6
研究了粗糙集理论在数据挖掘中的应用,提出了一种基于粗糙集理论的数据挖掘算法.首先对信息系统的数据加工泛化,构造其二进制可辨矩阵.对矩阵进行化简得到属性约简并生成规则.最后,结合银行申请信用卡的实例,利用上述方法进行数据挖掘,消去冗余属性,抽取决策规则. 相似文献
14.
分类发现是数据挖掘的一个重要任务,Bayes统计推断是数据挖掘中数据分类的主要方法也是其基础之一。建立在粗糙集基础上。利用概率工具以及Expectation-Maximization思想(RUBIN算法)和Bayes统计推断规则的结合提出了分类的方法和步骤,并给出实例分析,重点是RUBIN算法和Bayes统计推断规则相结合在粗糙集上建立的决策方法。 相似文献
15.
基于模糊集和粗糙集的关联规则挖掘策略 总被引:3,自引:0,他引:3
提出了一种对原始数据先进行模糊聚类,再提取规则的基于模糊集和粗糙集技术的关联规则挖掘策略,可以在一定程度内减少噪声数据的干扰,消除数据对象中的冗余属性,有利于提高规则挖掘的有效性. 相似文献
16.
基于粗糙集理论的复方拆方研究 总被引:2,自引:0,他引:2
粗糙集理论是一种新的处理模糊和不确定性知识的数学工具,其主要思想就是在保持分类能力不变的前提下,通过知识约简,导出问题的决策和分类规则.因此,将粗糙集理论引入复方拆方,将有助于从复方配伍的动态性和不确定性的本质上去深入而确切地开展方剂配伍规律的量化研究,从而改变药味与药效之间的分离状态. 相似文献
17.
用于粗糙集约简的并行算法 总被引:1,自引:0,他引:1
通过对数据挖掘粗糙集约简算法的研究, 提出一种基于区分能力指数的信息系统数据划分思想. 先将系统按属性区分能力分成若干子表, 再由子表的约简求原系统的约简, 这种思想较好地简化了布尔函数的化简过程. 根据该思想设计了一个属性约简并行算法, 并利用Petri网模拟工具CPN Tools对算法的负载平衡进行了分析, 通过实验与相应的串行算法在时间上进行了对比, 实验结果显示, 该算法对于对象较多的大规模系统具有较高的效率. 相似文献
18.
Rough集规则知识获取研究中的不一致性问题 总被引:5,自引:0,他引:5
讨论数据挖掘问题 ,即从原始数据中构造决策规则。重点考虑不一致情况下的规则知识获取问题 ,即从包含不一致信息的数据中获取得到缺省规则 ,并研究在不一致条件下的决策规则选择策略 ,使之能够在不完全、不一致的条件下进行推理。同时将在 Skowron缺省规则获取算法的基础上 ,根据对不一致性的分析 ,提出从包含不一致信息的决策表中获取缺省规则 ,并能够对任意待识样本进行处理的方法。 相似文献
19.
利用Rough集理论中关于等价类的概念,提出了单维布尔关联规则问题挖掘算法,考虑到关联规则设定单一最小支持度阈值的局限性,提出使用多个最小支持度的办法进行频繁项集的发现,利用兴趣度对单维布尔关联规则进行评价. 相似文献
20.
一种基于粗糙集的粗糙神经网络构造方法 总被引:3,自引:0,他引:3
针对神经网络中各神经元和权不能赋予明确的物理意义 ,提出了一种基于粗糙集的神经网络模型 .该模型利用粗糙集理论数值分析的能力 ,首先从给定的数据集中抽取出规则 ,然后根据这些规则构造神经网络隐含层的神经元个数 ,从而确定粗糙神经网络的初始拓扑结构 .同时 ,将输入映射到输出子空间 ,并在这个子空间上用神经网络进行逼近 ,由此得到一种可理解性好、收敛速度快的神经网络模型 .实验结果表明 ,该模型能够较好地处理神经网络拓扑结构、训练样本的大小、样本质量等对神经网络的精度及泛化能力有直接影响的问题 ,在大大缩短训练时间的同时 ,它的预测精度可达 96 4 % ,较同条件下径向基函数神经网络模型的精度高 3 6 % . 相似文献