基于粗糙集的证据理论及其在网络证据融合中的应用

发布时间：2018-03-05 00:29

本文选题：模糊聚类算法　切入点：D-S证据理论　出处：《山东师范大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年来伴随着网络的遍及,网络犯罪率的增长速度十分惊人,而在计算机安全领域有这样一个全新的分支——网络取证,受到了越来越多的关注。对于网络犯罪案件,需要采用有效的工具和手段进行网络取证分析。然而在证据融合的过程中,经常会遇到网络证据来源广、数量过多,证据信息具有不确定性等问题。本文针对网络证据的复杂性问题提出了一种反向遗传算法优化的模糊C均值算法,用于改善对网络证据进行数据预处理的效果;并提出了基于可变粒度粗糙集的证据理论融合方法,以及将该方法应用到网络证据的融合处理中。本文所做工作主要是以下这三个方面的内容:(1)研究模糊聚类,提出一种反向遗传算法优化的模糊C均值算法。该算法主要是考虑了遗传算法本身易早熟的缺陷对于模糊C均值算法的聚类结果会产生不良影响,因此通过引入反向学习机制,构造一种反向遗传算法,并应用于对模糊C均值算法的改进中。实验最终结果表明,该算法能够有效地提高聚类准确率,并加快整个算法的迭代效率。(2)研究粗糙集理论和证据理论之间的联系,提出一种基于可变粒度粗糙集的证据理论融合方法。该方法深入剖析了粒度计算视角下,粗糙集理论与D-S证据理论二者之间的联系,利用粒度空间即证据空间距离的计算判断证据之间是否冲突,并构造了一个基于可变粒度粗糙集的定性融合函数。通过实例计算证明,该方法能有效地处理多源证据空间的证据融合问题。(3)设计基于粗糙集-证据理论的网络证据融合模型,实现对网络通信数据流的分析处理。将结合了两种理论优势的改进证据融合方法应用到网络证据融合模型。在该系统中,首先是利用反向遗传算法优化的模糊C均值算法对采集到的数量大、结构复杂的原始数据进行聚类预处理,去除冗余;然后利用基于可变粒度粗糙集的证据理论融合方法对证据进行融合操作,获取一个一致的合理有效的证据融合结果,并生成取证报告。结果证实,本文提出的方法能够提高系统在网络证据融合方面的有效性。
[Abstract]:In recent years, with the pervading of the network, the growth rate of the network crime rate is very astonishing, and in the field of computer security, there is such a new branch-network forensics, which has received more and more attention. It is necessary to use effective tools and means for network forensics analysis. However, in the process of evidence fusion, we often encounter a large number of online evidence from a wide range of sources. For the complexity of network evidence, this paper presents a fuzzy C-means algorithm optimized by reverse genetic algorithm, which is used to improve the effect of data preprocessing on network evidence. A method of evidence theory fusion based on variable granularity rough set is proposed, and the method is applied to the fusion of network evidence. The main work of this paper is to study fuzzy clustering in the following three aspects: 1). This paper presents a fuzzy C-means algorithm optimized by reverse genetic algorithm. This algorithm mainly considers the disadvantage of precocity of genetic algorithm to the clustering result of fuzzy C-means algorithm, so the reverse learning mechanism is introduced. A reverse genetic algorithm is constructed and applied to the improvement of fuzzy C-means algorithm. The experimental results show that the algorithm can effectively improve the clustering accuracy. And accelerate the iteration efficiency of the whole algorithm. 2) study the relation between rough set theory and evidence theory, and propose a method of evidence theory fusion based on variable granularity rough set. The relation between rough set theory and D-S evidence theory. A qualitative fusion function based on variable granularity rough set is constructed. It is proved that the method can effectively deal with the evidence fusion problem in multi-source evidence space. The improved evidence fusion method, which combines two theoretical advantages, is applied to the network evidence fusion model. Firstly, the fuzzy C-means algorithm optimized by reverse genetic algorithm is used to preprocess the collected raw data with large quantity and complex structure, so as to remove the redundancy. Then the evidence theory fusion method based on variable granularity rough set is used to fuse the evidence to obtain a consistent, reasonable and effective evidence fusion result, and the evidence report is generated. The method proposed in this paper can improve the effectiveness of the system in the network evidence fusion.
【学位授予单位】：山东师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;D925.2

【相似文献】