协同聚类关键技术研究
发布时间：2023-04-17 03:34
本文中,我们研究了协同聚类,并将相关概念与信息安全中的聚类分析联系起来;在这个问题中,我们关注于纷繁复杂的网络攻击时代中,随着数据的数量和复杂性不断增长所带来的数据安全与隐私问题。在应用需求的推动下,我们引入了协同聚类框架,该框架能够为信息安全中的数据挖掘应用进行大型分布式数据库建模和网络建模。协同聚类符合信息安全中的数据挖掘需求主要体现在两个方面:首先是协同聚类能通过使用信息颗粒保证隐私,同时允许使用原型进行协同任务;其次,在面向具有高维大数据集和表示被监控对象行为的多个特征时,为算法提供可扩展性,这反过来不仅增加了学习正常行为问题的复杂性,而且还可能给聚类分析带来严重错误。然而,诸如协同模糊聚类、协同自组织映射和协同生成式拓补映射等协同聚类方法存在需要输入参数来决定协同信息影响的问题,这些参数对聚类结果又很大的影响,因此不能被忽视。我们提出了一种协同聚类框架,该框架使用粒子群优化来最小化聚类的熵,以寻找最佳聚类中心。此外,它使用粒子矢量位置更新来确定协同信息的重要性,从而消除了对用户输入参数的依赖。被称为粒子子圈的框架结合了来自几种聚类算法的信息,从而部分解决了选择正确聚类方法的问...
【文章页数】：102 页
【学位级别】：硕士
【文章目录】：
摘要
Abstract
LIST OF SYMBOLS
CHAPTER1:INTRODUCTION
1.1 OVERVIEW
1.2 CLUSTERING IN INFORMATION SECURITY
1.3 CHALLENGES
1.4 APPLICATIONS
1.5 CONTRIBUTION OF THIS THESIS
CHAPTER2:LITERATURE REVIEW
2.1 CHALLENGES IN TRADITIONAL CLUSTER ANALYSIS
2.2 DISTANCE AND SIMILARITY MEASURES
2.2.1 Distance measures
2.2.2 Similarity measures
2.3 PROTOTYPE-BASED CLUSTERING ALGORITHMS
2.3.1 K-Means
2.3.2 Fuzzy C-Means
2.3.3 Gaussian Mixture Models
2.3.4 Affinity Propagation Clustering
2.4 CLUSTER VALIDITY INDEXES
2.4.1 Internal Validity Indexes
2.4.2 External Validity Indexes
2.5 CHAPTER CONCLUSION
CHAPTER3:COLLABORATIVE CLUSTERING
3.1 CHALLENGES AND MODERN CLUSTER ANALYSIS
3.2 COLLABORATION SCHEMES
3.3 STATE OF THE ART IN COLLABORATIVE CLUSTERING
3.3.1 Collaborative Fuzzy c-means clustering
3.3.2 Prototype based collaborative algorithms
3.4 CHAPTER CONCLUSION
CHAPTER4:PARTICLE SUBSWARMS COLLABORATIVE CLUSTERING
4.1 THE COLLABORATIVE FUZZY CLUSTERING AND ITS CHALLENGES
4.2 DEFINITIONS
4.3 THE FRAMEWORK FUNDAMENTALS
4.3.1 Fitness Function
4.3.2 Particle Position Update
4.3.3 Stopping Criteria
4.4 THE DESIGN AND COMPLEXITY ANALYSIS
4.4.1 Collaboration with crisp clustering
4.4.2 Collaboration with fuzzy clustering
4.5 THE EXPERIMENTAL RESULTS
4.5.1 Crisp clustering results
4.5.2 Fuzzy clustering results
4.5.3 Comparison with other frameworks
4.6 CHAPTER CONCLUSION
CHAPTER5:CONCLUSION
5.1 PRIMARY FINDINGS
5.2 LIMITATIONS OF PSSCC
5.3 RECOMMENDATIONS FOR FUTURE WORK
ACKNOWLEDGEMENTS
REFERENCE
APPENDIX A:DATA SETS AND IMPLEMENTATIONS
A.1 EXPERIMENTAL DATA SETS
A.2 IMPLEMENTATIONS
APPENDIX B:CONVERGENCE AND PROOFS
B.1 CONVERGENCE OF K-MEANS
B.2 CONVERGENCE OF FUZZY C-MEANS
B.2.1 Expectation step
B.2.2 Maximization step
B.3 CONVERGENCE OF GAUSSIAN MIXTURE MODELS
B.3.1 Updating the mixing coefficients
B.3.2 Updating the centers of the clusters
B.3.3 Updating the covariance matrices
B.3.4 Corollary:EM algorithm and Gaussian Mixtures
本文编号：3792481
【文章页数】：102 页
【学位级别】：硕士
【文章目录】：
摘要
Abstract
LIST OF SYMBOLS
CHAPTER1:INTRODUCTION
1.1 OVERVIEW
1.2 CLUSTERING IN INFORMATION SECURITY
1.3 CHALLENGES
1.4 APPLICATIONS
1.5 CONTRIBUTION OF THIS THESIS
CHAPTER2:LITERATURE REVIEW
2.1 CHALLENGES IN TRADITIONAL CLUSTER ANALYSIS
2.2 DISTANCE AND SIMILARITY MEASURES
2.2.1 Distance measures
2.2.2 Similarity measures
2.3 PROTOTYPE-BASED CLUSTERING ALGORITHMS
2.3.1 K-Means
2.3.2 Fuzzy C-Means
2.3.3 Gaussian Mixture Models
2.3.4 Affinity Propagation Clustering
2.4 CLUSTER VALIDITY INDEXES
2.4.1 Internal Validity Indexes
2.4.2 External Validity Indexes
2.5 CHAPTER CONCLUSION
CHAPTER3:COLLABORATIVE CLUSTERING
3.1 CHALLENGES AND MODERN CLUSTER ANALYSIS
3.2 COLLABORATION SCHEMES
3.3 STATE OF THE ART IN COLLABORATIVE CLUSTERING
3.3.1 Collaborative Fuzzy c-means clustering
3.3.2 Prototype based collaborative algorithms
3.4 CHAPTER CONCLUSION
CHAPTER4:PARTICLE SUBSWARMS COLLABORATIVE CLUSTERING
4.1 THE COLLABORATIVE FUZZY CLUSTERING AND ITS CHALLENGES
4.2 DEFINITIONS
4.3 THE FRAMEWORK FUNDAMENTALS
4.3.1 Fitness Function
4.3.2 Particle Position Update
4.3.3 Stopping Criteria
4.4 THE DESIGN AND COMPLEXITY ANALYSIS
4.4.1 Collaboration with crisp clustering
4.4.2 Collaboration with fuzzy clustering
4.5 THE EXPERIMENTAL RESULTS
4.5.1 Crisp clustering results
4.5.2 Fuzzy clustering results
4.5.3 Comparison with other frameworks
4.6 CHAPTER CONCLUSION
CHAPTER5:CONCLUSION
5.1 PRIMARY FINDINGS
5.2 LIMITATIONS OF PSSCC
5.3 RECOMMENDATIONS FOR FUTURE WORK
ACKNOWLEDGEMENTS
REFERENCE
APPENDIX A:DATA SETS AND IMPLEMENTATIONS
A.1 EXPERIMENTAL DATA SETS
A.2 IMPLEMENTATIONS
APPENDIX B:CONVERGENCE AND PROOFS
B.1 CONVERGENCE OF K-MEANS
B.2 CONVERGENCE OF FUZZY C-MEANS
B.2.1 Expectation step
B.2.2 Maximization step
B.3 CONVERGENCE OF GAUSSIAN MIXTURE MODELS
B.3.1 Updating the mixing coefficients
B.3.2 Updating the centers of the clusters
B.3.3 Updating the covariance matrices
B.3.4 Corollary:EM algorithm and Gaussian Mixtures
本文编号：3792481
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/3792481.html
