粒子群优化的半监督入侵检测算法

为解决无监督入侵检测算法检测率低,有监督的入侵检测不能有效的检测未知攻击的问题,提出了一种粒子群优化的半监督入侵检测算法,算法对少量的约束信息进行基于密度的扩展获得潜在约束得到聚类模型,以此指导未标记数据聚类,对仍没有确定类别的未标示数据使用粒子群优化的K均值算法进行聚类实现对异常的检测。改进的算法检测率达到83.7%,误报率减少至3.13%,总体效果优于无监督和有监督学习的入侵检测算法。     

一种基于半监督学习的入侵检测算法

针对无监督的入侵检测检测效率较低,而有监督入侵检测算法不能有效的检测异常攻击,提出一种半监督学习的入侵检测算法,新算法先用有标记数据进行初始聚类,然后利用初始聚类指导未标记数据聚类,最后使用K近邻算法对仍没有确定类别的未标示数据对异常进行检测,结果表明,改进后算法的效果优于无监督和有监督学习的入侵检测算法。     

基于PCA和LOGIT模型的网络入侵检测方法

针对高维度网络数据进行多类攻击行为检测的需求,提出一种基于PCA和Logit模型的网络入侵检测方法,通过PCA对网络数据降维、简化数据集和多项式Logit模型实现对不同攻击行为的识别并判别其类型。实验结果表明,方法在网络入侵识别上具有较理想的效果。     

基于遗传算法的免疫网络设计

从人体免疫网络抵御抗原与计算机网络抵御入侵的相似性,提出一个基于遗传算法的免疫网络并应用于分布式入侵检测系统.将一个基于每秒网络数据量为参数的时间序列作为检测子的抗体表达形式,在单个网段检测网络数据,通过各个网段之间的抗体交流,实现抗体的优化.在抗体的进化过程中,提出了较好地评价抗体适应度的评估函数,为优化抗体的遗传算法的实现打下了基础.     

基于人工免疫网络的在线核聚类算法

针对核聚类算法与免疫网络聚类算法的不足,将免疫网络机制与核理沦相结合,通过用核距离函数代替欧拉距离函数,设置反映抗体识别抗原数量的权重和引入人工免疫网络机制等解决样本类边界模糊和类间数据密度分布不均匀问题.最后通过人工数据集和lRIS数据集上的仿真试验,验证了算法的有效性.     

基于聚类算法的钻孔瞬变电磁视电阻率立体成像方法

煤矿井下地球物理水害超前探测要求探测点20 m范围内不得有积水和金属物体,传统电磁法超前探测技术已不能满足要求,钻孔瞬变电磁法通过将收发装置送入掘进工作面前方的钻孔中进行探测,既远离了巷道中的各种干扰,又提高了隐蔽致灾水体的探测精度。为解决该方法对钻孔径向异常体的准确定位解释难题,通过三维正演总结了其水平分量异常响应特征,提出了异常体象限确定准则,研究了根据水平分量幅值和异常象限综合求取异常体工具面角的计算方法。将由垂直分量计算得到的每一个视电阻率视为独立异常体,基于K-means聚类算法对相应的水平分量异常曲线特征值进行二分类,实现了全数据集的视电阻率象限自动划分,结合异常工具面角算法研究得出钻孔瞬变电磁视电阻率立体成像方法。最后计算了三维数值模型的立体成像结果,对钻孔径向的小规模低阻异常体取得了良好效果。结果表明:基于K-means聚类算法的钻孔瞬变电磁视电阻率立体成像方法是地球物理与机器学习的有机结合,该方法能够为井下掘进工作面隐伏水害超前探测精细解释提供技术支撑。     

“21世纪海上丝绸之路”有向复杂航运网络抗毁性研究

为研究“21世纪海上丝绸之路”海运航线网络特征,运用有向加权复杂网络方法,从拓扑结构和航运网络性能两个角度,设计基于港口/航线的仿真体系,进而探讨海运航线网络特征及其抗毁性。结果表明:①“海上丝路”航运有向加权网络节点强度分布近似满足幂指数分布,同时具有较小的平均最短距离和较高的加权聚类系数,边权分布不均衡,少数航线承担了该区域主要班次,航线趋于集中化。②网络表现出随机攻击模式下的抗毁性和蓄意攻击下的脆弱性,其中基于节点的蓄意攻击或边的随机攻击对网络的破坏性最大。蓄意攻击下,节点强度对网络鲁棒性影响较为明显,加权介数对度权指标影响较为明显;随机攻击下,边失效对网络鲁棒性和加权网络效率影响较大,不同攻击模式对抗毁性测度影响差别较大。     

基于多传感器融合式航迹起始算法

提出了一种基于聚类分析和Kalman 滤波相结合的多传感器航迹起始算法.根据多传感器同一时刻对同一目标的观测值在空间呈团状的特征,运用聚类的方法解决数据融合问题.采用一种改进的粒子群(PSO)优化算法对多传感器观测数据进行聚类,结合聚类中心和目标预测值,应用Kalman滤波器估计目标状态,从而实现航迹起始.实验结果表明,该方法有效.     

基于行为聚类算法的土地利用聚类适宜性分析研究

为了在较少地类的基础上,深入研究土地利用变化过程,可把较多的地类合并成较少的具有重要变化特征的地类.本文运用基于行为聚类的方法,将净增加和净增加的地类合并或净减少和净减少的地类合并,但不能将净增加和净减少的地类合并.主要分为3个阶段实现:第1个阶段是完全不变阶段,聚类结果是形成综合的完全不变地类;第2个阶段是完全减少和完全增加阶段,该阶段将完全减少地类和完全增加地类分别进行合并;第3个阶段是转变阶段,该阶段计算了每一对有相同净变化方向地类的相互转变信息.基于行为聚类的算法在每一步的计算过程中都会保留净变化面积为常数,转变变化面积保留最大化.本文对3个聚类原则进行了数学证明,定义了6种聚类行为,以云南省土地利用分类体系为例,从面积变化和强度变化2个角度阐述了行为聚类方法的可行性和优势性.最后,与刘纪远等建立的土地利用遥感监测分类系统进行比较,结果表明:行为聚类算法聚类到9种类型时,类型总变化面积达到5.10%,比原始类型的总变化面积下降了0.06%;而基于遥感监测分类系统的6个一级分类将总变化减少至4.7%,与聚类算法比较,总变化面积减少了0.4%.实践证明,行为聚类的算法可更好地保留土地利用动态变化信息,证明了该聚类方法的有效性.     

空间聚类有效性评价方法对比与研究

空间聚类是当前地球信息科学与计算机科学领域共同关注的热点问题之一,常用来揭示空间数据分布规律以及发现空间数据异常。空间聚类有效性评价即对空间聚类结果进行定量、客观的评判,对于在实际应用中针对不同数据集选取最优的空间聚类算法以及确定最佳的聚类参数具有重要意义。首先选取并编程实现了数种空间聚类有效性评价方法,包括聚类中心的距离矩阵、距离方差、改进Hubert's统计、Davies-Bouldin、Calinski-Harabasz和基于信息论的空间聚类有效性评价方法等,同时提出了顾及簇间分离度和簇内紧凑度的空间聚类有效性评价方法,并通过试验分析验证了其可行性及有效性。然后在K-Means法对数据集进行聚类的基础上,对比研究了前述聚类有效性评价方法的特性及优缺点。     

浮动车轨迹数据聚类的有向密度方法

为了充分挖掘浮动车轨迹数据的潜在特性,本文在OPTICS空间密度聚类算法基础上,提出了一种有向密度的快速聚类方法(D-OPTICS)。该方法通过扇形空间邻域计算其有向密度信息,并基于方向信息约束其密度可连通性,通过有向可达距离曲线生成数据基本簇,最后,通过空间网格及类簇聚合等优化方法,实现其大规模浮动车轨迹数据的快速聚类处理。通过有向时空数据的聚类分析,发现浮动车轨迹的时空分布特性,以提取复杂路网的结构信息。本文以福州市大规模浮动车轨迹数据,对D-OPTICS进行了系统实验,分析表明,该算法可实现浮动车轨迹数据的快速有向密度聚类分析,有助于挖掘发现时空轨迹数据的分布规律,且基于聚类结果提取了福州市区复杂路网的有向拓扑结构图。同时,与DBSCAN及OPTICS等传统的密度聚类算法进行性能对比,实验表明,D-OPTICS算法能更好地支持大规模浮动车轨迹数据的处理要求。     

基于GIS的单元簇遗传建模及其在区域矿产预测中的应用

论述一种GIS支持下的基于"单元簇”概念和遗传建模的多源地学信息综合分析方法及在多目标矿产预测中的应用.在以往矿产定量预测中,根据随机抽样思想进行单元划分,对空间信息利用不够充分.用单元的空间组合("单元簇”)代替单元作为定量类比的基本单位,从而能较好实现地质变量值及其所反映的局部地质空间结构的定量类比;将单元作为GIS区图元,利用GIS空间分析功能实现对单元及单元簇的管理和操作;以已知样品判对率最大为适应值计算准则,用遗传算法建立线性判别函数,可快速有效地实现多目标矿产定位预测.以新疆康古尔塔格地区金矿预测实例说明了其应用效果.     

���������Ż�RBF�����ڶ�̬����Ԥ���е�Ӧ��

?????????????????????????????????????????????????RBF?????????????????????CIOA????RBF??????????????????????????????CIOA?????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????RBF???????????????????????У???Ч????????????????????     

Spatial-temporal Distribution Characteristics of Global Seismic Clusters and Associated Spatial Factors

Earthquakes exhibit clear clustering on the earth. It is important to explore the spatial-temporal characteristics of seismicity clusters and their spatial heterogeneity. We analyze effects of plate space, tectonic style, and their interaction on characteristic of cluster.Based on data of earthquakes not less than moment magnitude(M_w) 5.6 from 1960 to 2014, this study used the spatial-temporal scan method to identify earthquake clusters. The results indicate that seismic spatial-temporal clusters can be classified into two types based on duration: persistent clusters and burst clusters. Finally, we analysed the spatial heterogeneity of the two types. The main conclusions are as follows: 1) Ninety percent of the persistent clusters last for 22-38 yr and show a high clustering likelihood;ninety percent of the burst clusters last for 1-1.78 yr and show a high relative risk. 2) The persistent clusters are mainly distributed in interplate zones, especially along the western margin of the Pacific Ocean. The burst clusters are distributed in both intraplate and interplate zones, slightly concentrated in the India-Eurasia interaction zone. 3) For the persistent type, plate interaction plays an important role in the distribution of the clusters’ likelihood and relative risk. In addition, the tectonic style further enhances the spatial heterogeneity. 4) For the burst type,neither plate activity nor tectonic style has an obvious effect on the distribution of the clusters’ likelihood and relative risk. Nevertheless,interaction between these two spatial factors enhances the spatial heterogeneity, especially in terms of relative risk.     

Application of different clustering approaches to hydroclimatological catchment regionalization in mountainous regions,a case study in Utah State

With respect to the different hydrological responses of catchments, even the adjacent ones, in mountainous regions, there are a great number of motivations for classifying them into homogeneous clusters. These motivations include prediction in ungauged basins (PUB), model parameterization, understanding the potential impact of environmental changes, transferring information from gauged catchments to the ungauged ones. The present study investigated the similarity of catchments through the hydro-climatological pure time-series of a 14-year period from 2001 to 2015. Data sets encompass more than 13,000 month-station streamflow, rainfall, and temperature data obtained from 27 catchments in Utah State as one of the eight mountainous states of the USA. The identification, analysis, and interpretation of homogeneous catchments were investigated by applying the four approaches of clustering, K-means, Ward, and SOM (Self-Organized Map) and a newly proposed Wavelet-Entropy-based (WE-SOM) clustering method. By using two clustering evaluation criteria, 3, 5, and 6 clusters were determined as the best numbers of clusters, depending on the method employed, where each cluster represents different hydro-climatological behaviors. Despite the absence of geographic characteristics in input data matrix, the results indicated a regionalization in agreement with topographic characteristics. Considering the dependency of the hydrological behavior of catchments on the physiographic field aspects and characteristics, WE-SOM method demonstrated a more acceptable performance, compared to the other three conventional clustering methods, by providing more clusters. WE-SOM appears to be a promising approach in catchment clustering. It preserves the topological structure of data which can, as a result, be proofed in a greater number of clusters by dividing data into higher numbers of distinct clusters with similar altitudes of catchments in each cluster. The results showed the aptitude of wavelets to quantify the time-based variability of temperature, rainfall and streamflow, in the way contributing to the regionalization of diverse catchments.     

利用空间聚集的贝叶斯网络评估手足口病发病风险

手足口病是一种常见的传染病,以往的研究表明该疾病与气象、环境和社会经济等因素相关联,其影响关系复杂,而疾病本身体现出较强的区域聚集性,采用普通的线性风险建模方法无法捕捉影响因素的复杂性及空间聚集性。因此,本文以山东省为例,在前人研究的基础上,提出了采用贝叶斯网络综合风险建模方法研究手足口病的发病风险与气象、土地利用、社会经济及空气污染等要素间的关系,并通过引入空间扫描统计聚集结果,将空间聚集引入到贝叶斯网络模型加强其空间推理功能,减少模型的偏差,提高评估的精度。结果表明,本文建立的手足口病空间贝叶斯网络风险模型具有较高的估计效果,引入的空间聚集性较好地融入到贝叶斯概率推理模型中,合理建立预测因子同手足口病发病风险之间的关系。通过对建模结果的解译,分析了手足口病的发病风险影响因素,特别是气候、社会经济及空气污染的影响。本文的空间贝叶斯建模方法及研究结果对手足口病暴发的防控预警具有重要的意义。     

Evolution process of urban spatial pattern in Hubei Province based on DMSP/OLS nighttime light data

This paper intends to investigate the urban spatial patterns of Hubei Province and its evolution from three different perspectives: urban nodes, urban connections and urban clusters. The research adopts nighttime light imagery of cities in Hubei Province, the viewpoint of ′point-axis-area′ in the ′point-axis system′ theory, and employs light index model, gravity model and social network analysis. The findings are as follows: 1) In terms of urban nodes, the urbanization process of Hubei has been carried out mainly on the basis of external expansion rather than internal increasing. The polarization trend of urban connection network is strengthening. 2) As for urban connections, the estimation of urban connections using light index model is capable of containing various actual flow, and the connections are getting increasingly closer. 3) In regard to urban groups, seven urban groups of varying sizes have formed. On that basis, three stable and relatively independent urban groups as the centers, namely Wuchang, Yichang and Xiangyang emerge as well. But the structures of ′Wuhan Metropolitan Area′, ′Yichang-Jingzhou-Jingmen City Group′ and ′Xiangyang-Shiyen-Suizhou City Group′, which are defined by local development strategy in Hubei Province, are different from the above three urban groups.     

Triterpene-Enriched Olive Extract as an Immunopotentiator in Black Sea Bream ( Acanthopagrus schlegelii)

Recently,the use of natural immunopotentiators,such as plant extracts,is predicted to offer high disease-prevention potential in aquaculture.However,few reports on the immunomodulatory impacts of olive extract(OE)on teleost are available.Therefore,the effects of dietary intake of triterpene-enriched OE on black sea bream(Acanthopagrus schlegelii)were investigated in this study.Our data showed that total blood cell counts,in vivo content of lysozyme,activities of antiprotease and myeloperoxidase,and contents of IL-4 and IL-6 were significantly up-regulated by dietary intake of triterpene-enriched OE.Additionally,the hampered immune response induced by cadmium exposure was significantly mitigated by the administration of OE as were indicated by partially or completely rescued immune-relating parameters.Furthermore,the expressions of immune-related genes encoding NF-κB inhibitor alpha(IkBα),tumor necrosis factorα(TNFα),cyclooxygenase-2(COX-2),and proto-oncogene protein c-fos(FOS)were found to be significantly up-regulated by the dietary intake of OE.In general,the results suggested that the dietary intake of triterpene-enriched OE has immune enhancing effect in black sea bream.Such effect may be realized by 1)increasing the total counts of diverse blood cells;2)activating nonspecific immune biomolecules;and 3)affecting signaling pathways such as IKK and ERK and subsequently inducing IL-4 and IL-6 biosyntheses.