基于离散度的决策树构造方法

在构造决策树的过程中,属性选择将影响到决策树的分类精度.对此,讨论了基于信息熵方法和WMR方法的局限性,提出了信息系统中条件属性集的离散度的概念.利用该概念在决策树构造过程中选择划分属性,设计了基于离散度的决策树构造算法DSD.DSD算法可以解决WMR方法在实际应用中的局限性.在UCI数据集上的实验表明,该方法构造的决策树精度与基于信息熵的方法相近,而时间复杂度则优于基于信息熵的方法.     

基于Rough集理论的Self集的构造和演化算法

提出了一种基于Rough集理论的Self集构造和演化算法。利用Rough集约简算法,对用户的安全访问行为的数据作规范化处理并进行约简,从中提取有效的最简规则,降低了安全数据的冗余,减轻了特征码构造的负担。使用Rough集上、下近似集原理,构造了上、下近似Self集,实现了Self的优化和扩展,有效地解决了Self集的自动演化问题。     

基于Rough集理论的Self集的构造和演化算法

提出了一种基于Rough集理论的Self集构造和演化算法。利用Rough集约简算法，对用户的安全访问行为的数据作规范化处理并进行约简，从中提取有效的最简规则，降低了安全数据的冗余，减轻了特征码构造的负担。使用Rough集上、下近似集原理，构造了上、下近似Self集，实现了Self的优化和扩展，有效地解决了Self集的自动演化问题。     

基于变异的检测器生成算法研究

为提高基于免疫的网络入侵检测系统中检测器的生成效率,减小计算量.对Forrest的否定选择算法(NSA)进行改进,提出候选检测器集的生成不再采用随机方式,而通过对两个数据集(一是已有的合格检测器集,二是自我数据集)进行变异来产生,即利用部分已有的检测结果反馈生成成熟检测器.改进算法提高了候选检测器成为成熟检测器的比率,实验结果表明了算法的有效性.     

实值n维混沌映射否定选择算法

针对传统的基于二进制的混沌否定选择算法在检测器生成阶段对混沌映射产生的混沌序列离散化生成的候选检测器,不利于知识和数据的分析,也会造成检测器集生成速度慢及检测效率低等问题,提出了基于实值的混沌否定选择算法.引入混沌理论,采用混沌特性更好的自映射构造n维混沌映射生成候选检测器中心点,改进了传统的检测器生成机制,更适合处理高维空间问题;对原有的V-detector算法进行了优化,通过定向移动与计算几何中心相结合的思想确定检测半径.旨在满足预期覆盖率条件下尽量使半径取值最大化,扩大检测器集的覆盖范围,减少检测器数量.实验结果表明,该算法提高了检测器集的生成速度和检测效率.     

一种新的人工免疫系统检测规则及其应用

为了提高检测器集的生成效率,在讨论人工免疫系统负选择模型的基础上,提出一种新的检测规则,即:编辑距离规则.在这种规则中,对于一个随机生成的字符串和Self集中的字符串,采用编辑距离度量它们之间的相似性.随后,利用这种检测规则给出一种新的检测器集生成算法,它要求利用Trie数据结构组织和存储Self集.最后,通过理论分析得出了使用该种算法的优越性.     

基于免疫的入侵检测方法研究

生物的免疫系统和计算机安全系统所面临及需要解决的问题十分类似．采用生物免疫思想的入侵检测技术可以结合异常检测和误用检测的优点．研究了基于免疫的入侵检测方法，对Self集的确定和有效检测器的生戍方法进行了研究和改进，基于反向选择机制提出了一种新的有效检测器生成算法．可以使用较少的有效检测器检测网络中的异常行为，从而提高了有效检测器生成和入侵检测的速度．通过与基于已有的有效检测器生成算法的系统进行比较，使用本文的方法构造的入侵检测系统速度更快．且有较高的准确性．     

计算机免疫识别规则的演化挖掘

识别器的构造是计算机免疫学的一个重要研究方向。提出一种构造识别器的演化算法,其核心思想是采用数据挖掘的方法从Self集和nonself集中挖掘识别规则,这些识别规则能够反映出Self和nonself的一些内在特征,因此有很好的识别能力。     

基于知识粗糙度的决策树生成方法

决策树是一种重要的数据分类方法,测试属性的选择直接影响到决策树中结点的个数和深度,本文提出了一种基于知识粗糙度的方法.通过比较我们发现:在决策树的构造上,粗集理论中知识粗糙度的方法计算量较小,构造的决策树比经典ID3算法简洁,并且具有较高的分类精度.     

一种采用混合检测器的入侵检测系统

传统的基于免疫的入侵检测系统采用低级别的二进制检测器,妨碍了有意义的知识提取,对Nonself空间的覆盖也不完备。对二进制Self集的确定和有效检测器的生成方法进行了改进,研究了实值否定选择算法,加入了实值检测器,构成混合检测器集合,在检测阶段对会话和数据包同时进行异常检测。实验结果ROC曲线表明有较高的检测率和较低的误报率。     

免疫算法在入侵检测模型中的应用研究

在对常见的免疫算法原理进行分析的基础上,采用阴性选择算法和r-连续位匹配算法,提出一种改进的免疫检测机制,建立一个新的入侵检测模型。新的模型主要采取三点措施:改进候选检测集的生成规则;降低检测器冗余;引入协同检测机制等。在入侵识别阶段,采用基于编辑距离的匹配规则,提高了检测效率。试验仿真表明,该模型可有效提高入侵检测系统的检测率,降低误警率。     

网络入侵检测系统自体集检测中的概率匹配高效寻优机制

针对自体集数据规模较大造成的时空上的巨大消耗而难以处理的问题,设计了基于人工免疫的网络入侵检测系统(NIDS)的自体集匹配机制。为提高入侵检测系统的检测效率,提出概率匹配高效寻优机制。首先证明了网络数据的相对集中性,通过计算平均查找长度(ASL)分析了概率匹配机制的有效性,并通过模拟实验验证了该机制的快速匹配效率,并且在一种基于自体集规模简约机制的新型人工免疫网络入侵检测系统上进行了工程应用,取得了较好的匹配效果。     

免疫检测器证据理论集成的机组复合故障诊断

针对机组复合故障诊断准确率较低的状况,基于免疫机理的人工免疫智能方法,构建对故障比较敏感的无量纲指标免疫检测器．采用自适应调节匹配阈值和从非己空间产生的候选检测器,能有效减少黑洞和边界不清晰．通过免疫编程优化策略获得最佳识别能力的新特征指标．利用证据理论对多类免疫检测器进行集成诊断,提炼出能直接应用于复合故障诊断的优秀无量纲免疫检测器,机组实验结果表明,所得免疫检测器能快速、准确地进行复合故障诊断．     

How commercial involvement affects open source projects: three case studies on issue reporting

Whereas most research on Internetware has focused on new technologies for keeping track of a changing Internet,little attention has been paid to the software development process.A large portion of the software running the Internet is open source software.Open source software is developed both by volunteers and commercial companies,often jointly.Companies get involved in open source projects for commercial reasons,and bring with them a commercial software development process.Thus,it is important to understand how commercial involvement affects the software development process of open source projects.This article presents case studies of three open source application servers that are being developed jointly by a volunteer community and one primary software company.We are interested in better understanding developer behavior,specifically task distribution and performance,based on whether the developer is an external contributor,e.g.,a volunteer working in their spare time,or a commercial developer from inside the primary backing company who is being paid for their time.To achieve this,we look at issue reporting as an example of commercial involvement in open source projects.In particular,we investigate the distribution of tasks among volunteers and commercial developers by studying the source of reported issues and quantify the task performance on user experience via the issue resolution speed.We construct measures based on historical records in issue tracking repositories.Our results show that,with intensified commercial involvement,the majority of issue reporting tasks would be undertaken by commercial developers,and issue resolution time would be reduced,implying a better user experience.We hope our methods and results provide practical insights for designing an efficient hybrid development process in the Internetware environment.     

Improved email spam detection model with negative selection algorithm and particle swarm optimization

The adaptive nature of unsolicited email by the use of huge mailing tools prompts the need for spam detection. Implementation of different spam detection methods based on machine learning techniques was proposed to solve the problem of numerous email spam ravaging the system. Previous algorithm used in email spam detection compares each email message with spam and non-spam data before generating detectors while our proposed system inspired by the artificial immune system model with the adaptive nature of negative selection algorithm uses special features to generate detectors to cover the spam space. To cope with the trend of email spam, a novel model that improves the random generation of a detector in negative selection algorithm (NSA) with the use of stochastic distribution to model the data point using particle swarm optimization (PSO) was implemented. Local outlier factor is introduced as the fitness function to determine the local best (Pbest) of the candidate detector that gives the optimum solution. Distance measure is employed to enhance the distinctiveness between the non-spam and spam candidate detector. The detector generation process was terminated when the expected spam coverage is reached. The theoretical analysis and the experimental result show that the detection rate of NSA–PSO is higher than the standard negative selection algorithm. Accuracy for 2000 generated detectors with threshold value of 0.4 was compared. Negative selection algorithm is 68.86% and the proposed hybrid negative selection algorithm with particle swarm optimization is 91.22%.     

An antigen space density based real-value negative selection algorithm

The negative selection algorithm (NSA) is an important detector generation algorithm for artificial immune systems. In high-dimensional space, antigens (data samples) distribute sparsely and unevenly, and most of them reside in low-dimensional subspaces. Therefore, traditional NSAs, which randomly generate detectors without considering the distribution of the antigens, cannot effectively distinguish them. To overcome this limitation, the antigen space density based real-value NSA (ASD-RNSA) is proposed in this paper. The ASD-RNSA contains two new processes. First, in order to improve detection efficiency, ASD-RNSA utilizes the antigen space density to calculate the low-dimensional subspaces where antigens are densely gathered and directly generate detectors in these subspaces. Second, to eliminate redundant detectors and prevent the algorithm from prematurely converging in high-dimensional space, ASD-RNSA suppresses candidate detectors that are recognized by other mature detectors and adopts an “antibody suppression rate” to replace the “expected coverage” as the termination condition. Experimental results show that ASD-RNSA achieves a better detection rate and has better generation quality than classical real-value NSAs.     

Using known nonself samples to improve negative selection algorithm

Negative selection algorithm is the core algorithm of artificial immune system. It only uses the self for training and generates detectors to detect abnormalities. Holes are feature space areas that the detector fails to cover, it is the root cause of the performance degradation of the negative selection algorithm. The conventional method generates a large number of detectors randomly to repair the holes, which is time-consuming and not effective. To alleviate the problem, we propose a V-Detector-KN algorithm in this paper. V-Detector is the abbreviation of the real-valued negative selection algorithm with Variable-sized Detectors, KN represents Known Nonself. The V-Detector-KN algorithm uses the known nonself as the candidate detector to further generate the detector based on the V-Detector randomly generated detector, so as to realize the repair of holes. Compared with the conventional method to randomly generate detectors to repair holes, our proposed V-Detector-KN method uses known nonself to repair holes, reducing the randomness and blindness of hole repair. Theoretical analysis shows that the detection rate of our algorithm is not lower than that of the conventional V-Detector algorithm. The results of experiment comparing with other 6 algorithms on 7 UCI data sets show the superiority of our proposed algorithm.

     

Intelligent multi-user detection using an artificial immune system

Artificial immune systems (AIS) are a kind of new computational intelligence methods which draw inspiration from the human immune system. In this study, we introduce an AIS-based optimization algorithm, called clonal selection algorithm, to solve the multi-user detection problem in code-division multiple-access communications system based on the maximum-likelihood decision rule. Through proportional cloning, hypermutation, clonal selection and clonal death, the new method performs a greedy search which reproduces individuals and selects their improved maturated progenies after the affinity maturation process. Theoretical analysis indicates that the clonal selection algorithm is suitable for solving the multi-user detection problem. Computer simulations show that the proposed approach outperforms some other approaches including two genetic algorithm-based detectors and the matched filters detector, and has the ability to find the most likely combinations.