采用信息增益率的混合入侵检测模型设计

针对现有混合入侵检测模型仅定性选取特征而导致检测精度较低的问题，同时为了充分结合误用检测模型和异常检测模型的优势，提出一种采用信息增益率的混合入侵检测模型.首先，利用信息增益率定量地选择特征子集，最大程度地保留样本信息；其次，采用余弦时变粒子群算法确定支持向量机参数构建误用检测模型，使其更好地平衡粒子在全局和局部的搜索能力，然后，选取灰狼算法确定单类支持向量机参数构建异常检测模型，以此来提高对最优参数的搜索效率和精细程度，综合提高混合入侵检测模型对攻击的检测效果；最后，通过两种数据集进行仿真实验，验证了所提混合入侵检测模型具有较好的检测性能.     

Unresolved controversies concerning psychopathy: Implications for clinical and forensic decision making.

At some point in their careers, clinicians who work or consult in forensic and correctional settings will almost certainly encounter individuals who exhibit psychopathic personality features. Because of the widespread use of this disorder to inform legal and clinical decision making, psychologists should be exceedingly familiar with the relevant research literature on this topic before venturing into these settings. This article reviews the empirical bases of several clinically relevant claims and assertions regarding psychopathy and concludes that many areas of research are decidedly more equivocal in their findings than is commonly perceived. Although there is much to be gained by assessing psychopathy in various contexts, clinicians need to be cautious about drawing overzealous and empirically questionable conclusions about an important disorder that also has great potential for abuse. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

网络安全事件的关联分析方法的比较研究

随着当前攻击手段和技术的日益复杂化,一次入侵事件往往需要多个步骤才能完成,这些步骤都是彼此相关的。但是传统的入侵检测集中于检测底层的入侵或异常,所检测到的结果也仅仅是一次完整入侵的一部分,所以不能将不同的报警结合起来以发现入侵的逻辑步骤或者入侵背后的攻击策略。关联分析技术将不同分析器上产生的报警进行融合与关联分析,极大地减少了报警的数量,降低了入侵检测的误报率,并且适当的减少了入侵检测的漏报率。文中在对网络安全事件关联分析方法的系统结构进行分析后,着重介绍了当前比较流行的几种网络安全事件关联分析方法,最后对各种方法进行了比较研究。     

渭河口拦门沙形成及消失条件

在分析了黄、渭河部分测站的水文资料之后，研究渭河口拦门沙形成及消失条件，为减轻渭河下游的淤积寻找方法。     

Critical incidents and chronic stressors at work: Their impact on forensic doctors.

Workers in medium- or high-risk professions are often confronted with critical incidents at the workplace. The impact of these acute stressors may be serious and enduring. Many workers also experience chronic job stressors, such as work overload or role conflicts. This study examined the frequently neglected relationship of acute and chronic stressors with self-reported health symptoms, such as posttraumatic responses, fatigue, and burnout. This association was investigated in a sample of forensic doctors in the Netherlands (N = 84). It was found that the more traumatic events the respondents experienced, the more problems they reported in coping with the traumatic events. Chronic job stressors were associated with posttraumatic responses (intrusions and avoidances) and with burnout and fatigue. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Risk evaluations for the courts: Is service quality a function of specialization?

Psychologists are often asked to testify in court. However, concerns have been expressed in both the legal and psychological literatures about the quality of these services. This article presents survey results comparing the practice patterns of generally licensed psychologists with those of specialist forensic diplomates in providing risk evaluations. Results suggest that general clinicians frequently perform these evaluations, but forensic diplomates use more modern risk instruments, are more aware of the scientific literature, and provide the court with more information about the scientific basis of their testimony. The implication of these results for legal professionals, psychologist training, and practice are discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

利用茚三酮或DFO加强显现热敏纸张上的潜在指纹

本文介绍一种新技术——利用茚三酮或DFO加强显现热敏纸张上的潜在指纹。大部分热敏纸张的热敏剂表面,使用DFO或茚三酮石油醚溶液(NPB)显影时将会变黑。显现的指纹和背景之间反差很小。新技术介绍了在指纹和热敏层之间,用丙酮水洗,可以去掉黑色污斑。其它新的化学试剂漂洗,显现的指纹有清晰的纹线和很高的反差。大量使用这种技术,可以使试剂优化,成本降低,短时间内能处理大量检材。工作试剂包括商用、不挥发、含氮的有机化合物,也使用象NPB这样的浸透性试剂。     

基于统计阈值的Snort规则集动态产生的设计与实现

Snort作为开源的入侵检测系统,利用定义的静态规则集合实现对网络的入侵事件的检测。本文分析入侵检测系统的基本原理和模型,阐述Snort入侵检测系统部署到网络时,其静态规则集的配置方法,根据统计流量阈值和告警频率阈值动态产生动态规则集的方法,改进并提高了部署Snort应有的灵活性。     

一种基于网络安全风险评估的入侵检测方法

针对传统的入侵检测系统存在报警数量大、误报率高等缺陷,提出了一种基于网络安全风险评估的入侵检测方法,该方法基于入侵检测结果,引入抗体浓度随入侵强度动态变化这一人工免疫理论的最新研究成果进行网络安全风险的计算,然后根据当前网络面临的实时安全风险动态设置报警策略。实验结果表明,该方法能够实时、定量地计算主机和网络所面临的风险,并极大地降低报警数量和误报率。     

一种基于有向二分图模型和贝叶斯网络的入侵检测方法

针对入侵检测中存在的非确定性推理问题,文章提出一种基于二分图模型和贝叶斯网络的入侵检测方法,该方法利用二分有向图模型表示入侵和相关特征属性之间的因果拓扑关系,利用训练数据中获取模型的概率参数,最后使用最大可能解释对转化后的推理问题进行推理,并通过限定入侵同时发生的数目来提高检测效率。实验表明,该方法具有较高的检测率和很好的鲁棒性。