聚类分析的HMM自律可信性评估
首发时间:2010-03-30
摘要:为实现自律计算系统的自律性定量评估, 用聚类分析方法对自律服务状态进行分类,提取有效状态建立HMM模型,并采用信息熵优化模型参数,实现对系统服务的自律可信性定量分析. 实验结果表明: 聚类分析降低了对随机状态的选取,可以准确地反映自我管理能力要素之间的逻辑关系和变化,提高了评估分析的准确性。
For information in English, please click here
HMM Based on Clustering Analysis for Autonomic-dependability Evaluation
Abstract:In order to achieve quantitative evaluation for autonomic computing system, a HMM method based on clustering analysis is proposed, which autonomic service state is classified by clustering analysis, and effective state is extracted into HMM. Then, information entropy is applied to optimize the model parameter. Quantitative analysis for atonomic-dependability of system service is completed. Experiment result shows that clustering analysis reduces selection for random state which improves the model ability to the recognition of target system. The model can accurately reflect the logic relation and transformation for self-management elements, that improves accuracy for evaluation analysis.
Keywords: Autonomic evaluation Autonomic-dependability Clustering analysis HMM
论文图表:
引用
No.4134551128212699****
同行评议
勘误表
聚类分析的HMM自律可信性评估
评论
全部评论0/1000