基于参数优化隐马尔可夫模型的网络取证证据融合方法
发布时间：2018-12-28 17:53
【摘要】：为了克服隐马尔可夫模型(hidden Markov model,HMM)在训练时波氏算法(Baum-Welch,B-W)易陷入局部最优解的不足,采用自适应遗传算法对其进行参数优化,设计了染色体编码方法和遗传操作方式。利用Viterbi算法选择最有可能的元证据序列,用疑似证据替换元证据回溯得到证据链。实验结果表明,自适应遗传算法优化的HMM具有更好的状态,采用Viterbi算法得到的证据链能够较精确地重现网络入侵的犯罪现场。
[Abstract]:In order to overcome the deficiency that the hidden Markov model (hidden Markov model,HMM) is prone to fall into the local optimal solution in training, the adaptive genetic algorithm (AGA) is used to optimize the parameters of the algorithm. Chromosome coding and genetic manipulation were designed. The Viterbi algorithm is used to select the most probable meta-evidence sequence, and the suspect evidence is replaced by the meta-evidence backtracking to obtain the evidence chain. The experimental results show that the HMM optimized by the adaptive genetic algorithm has a better state, and the evidence chain obtained by using the Viterbi algorithm can accurately reproduce the crime scene of the network intrusion.
【作者单位】： 山东师范大学信息科学与工程学院;山东省分布式计算机软件新技术重点实验室;山东财经大学数学与数量经济学院;
【基金】：国家自然科学基金资助项目(61373148) 国家社科基金资助项目(12BXW040) 国家教育部人文社科基金资助项目(14YJC860042) 山东省自然科学基金资助项目(ZR2012FM038,ZR2014FL010) 山东省优秀中青年科学家奖励基金资助项目(BS2013DX033) 山东省高等学校科技计划资助项目(J15LN02)
【分类号】：D918.2;TP18
，
本文编号：2394250
[Abstract]:In order to overcome the deficiency that the hidden Markov model (hidden Markov model,HMM) is prone to fall into the local optimal solution in training, the adaptive genetic algorithm (AGA) is used to optimize the parameters of the algorithm. Chromosome coding and genetic manipulation were designed. The Viterbi algorithm is used to select the most probable meta-evidence sequence, and the suspect evidence is replaced by the meta-evidence backtracking to obtain the evidence chain. The experimental results show that the HMM optimized by the adaptive genetic algorithm has a better state, and the evidence chain obtained by using the Viterbi algorithm can accurately reproduce the crime scene of the network intrusion.
【作者单位】： 山东师范大学信息科学与工程学院;山东省分布式计算机软件新技术重点实验室;山东财经大学数学与数量经济学院;
【基金】：国家自然科学基金资助项目(61373148) 国家社科基金资助项目(12BXW040) 国家教育部人文社科基金资助项目(14YJC860042) 山东省自然科学基金资助项目(ZR2012FM038,ZR2014FL010) 山东省优秀中青年科学家奖励基金资助项目(BS2013DX033) 山东省高等学校科技计划资助项目(J15LN02)
【分类号】：D918.2;TP18
，
本文编号：2394250
本文链接：https://www.wllwen.com/falvlunwen/fanzuizhian/2394250.html
