基于标记树的WEB页面净化技术研究

根据Web页面标记建立标记树,通过分析,保留有用信息的标记子树,达到获取页面主要内容,净化页面的效果.     

决策树技术在汽车销售中的应用研究

利用sq l server analysis service建立轿车市场营销分析的数据仓库模型,在此基础之上,利用决策树分类技术,对汽车贸易企业的轿车销售历史数据进行分析,得出客户数据的分类规则,分析出潜在客户共同的特点;针对这些特点,采取相应营销决策,提高营销的针对性和成功率。     

基于双因素的可转换债券定价模型

将信用风险、利率期限结构引入到二叉树定价模型中,建立以股价和利率为标的变量的二叉树定价模型.在给定的边界条件和具体参数下,对我国的可转换债券进行定价研究,并用市场数据对模型做了实证检验.对比标的股价二叉树定价模型的计算结果,发现基于双因素的可转债定价模型的计算精确度较高,同时,计算得到的理论价值的走势跟市场价格的走势较为相似.     

基于离散化修正模糊算子的模糊故障树新析

针对不同类型隶属函数的模糊概率耦合计算的模糊故障树分析瓶颈问题尚未解决,提出了一种新的离散化方法,该算法通过先分割概率后规划隶属度的方式将模糊概率的隶属函数离散化,并由扩展原理修正与、或门模糊算子得到一组新的模糊算子,用新算子进行模糊故障树实例分析表明,新方法不但较好地解决了上述难点,并具有一定理论和工程实践价值。     

基于隐Markov树的设备状态综合诊断模型

为了充分利用小波系数之间的统计相依性以更有效地诊断设备状态,提出了一种基于隐Markov树(HMT)的综合诊断模型。首先通过主成分分析将来自多个传感器的信号转换为主成分,求出各主成分对应的频谱,然后通过比较对已训练的各HMT模型的适应度,运用Bayes决策融合法则得到设备状态综合诊断决策。为了克服HMT模型存在的计算溢出困难,采用尺度变换对EM算法进行了改进。通过两个实例验证了该综合诊断模型具有较高的诊断准确率。     

基于二元语义信息处理的最大树聚类方法

针对解决具有语言评价信息的多指标聚类分析问题,提出了一种基于二元语义信息处理的最大树聚类方法。该方法依据传统的最大树聚类分析方法的基本思想,采用二元语义概念对语言评价信息进行处理和运算,并对聚类对象进行分类,具有概念清晰、计算简单、信息处理较为精确的特点。最后,通过一个算例说明了所给出的聚类方法。     

可靠性仿真在模糊故障诊断中的应用

针对模糊故障诊断中传统模糊关系矩阵确定方法不能动态适应具体设备运行状态变化等缺点,提出了基于故障树的可靠性仿真方法,用于确定模糊关系矩阵。该方法首先利用元器件的失效分布数据和设备深层原理性知识建立故障树,进行可靠性仿真,然后根据仿真结果和具体设备的历史故障信息动态计算模糊关系矩阵。给出了该方法的具体应用步骤。实际诊断结果表明,该方法可以有效地提高模糊故障诊断的准确性与针对性。     

Magnetic properties of highroad-side pine tree leaves in Beijing and their environmental significance

This is a report on magnetic properties of highroad-side tree (Pinus pumila Regel) leaves collected along an expressway linking Beijing City and the Capital International Airport and further focus on their environmental contributions. A series of rock magnetic experiments show that the primary mag-netic mineral of leaf samples was identified to be magnetite,in the pseudo-single domain (PSD) grain size range 0.2―5.0 μm. On the other hand,magnet-ite concentration and grain size in leaves are ascer-tained to decrease with increasing of sampling dis-tance to highroad asphalt surface,suggesting that high magnetic response to traffic pollution is localized within a distance of some two meters away from highroad asphalt surface. Although magnetic suscep-tibility is ordinarily regarded as a simple,rapid and low-cost method for monitoring traffic pollution,saturation isothermal remanent magnetization (SIRM) can be treated as a valid proxy for monitoring air par-ticulate matter (PM) when samples are magnetically weak. It is believed that a synthetic rock magnetic study is an effective method for determining concen-tration and grain size of ferromagnets in the atmos-pheric PM,and then it should be a rapid and feasible technique for monitoring atmospheric pollution.     

Merkle Tree Digital Signature and Trusted Computing Platform

0 IntroductionDigital signatureis a key technique to provide datainteg-rity andsource authenticationcapabilities to enhance da-ta trustworthiness in trusted computing platform. Most cur-rent digital signature schemes are based on complex mathe-matical problems , such as integer factoring, discrete log-arithm,and discrete logarithm on elliptic curves . The hard-ness of these underlying mathematical problems is really un-certain at present . Merkle treeis a method used for authenti-cation and di…     

An Improved Model of Attack Probability Prediction System

This paper presents a novel probability generation algorithm to predict attacks from an insider who exploits known system vulnerabilities through executing authorized operations. It is different from most intrusion detection systems （IDSs） because these IDSs are inefficient to resolve threat from authorized insiders. To deter cracker activities, this paper introduces an improved structure of augmented attack tree and a notion of ＂minimal attack tree＂, and proposes a new generation algorithm of minimal attack tree. We can provide a quantitative approach to help system administrators make sound decision.