一种基于时序行为的流过程协同重构算法
过程流数据具有实时性、连续性和时序性等特征,使得传统过程挖掘算法难以发现隐含信息和演化过程。针对流过程模型的动态演化和重构要求,提出了一种基于时序行为分析的自适应混合启发式协同优化算法。首先定义演化流过程模型,基于日志活动间的隐含依赖关系改进过程逻辑的启发式挖掘规则,然后定义基于时序行为的老化因子,并引入高斯变异的多种群协作的自适应策略,改进粒子群优化算法的全局和局部精确寻优能力,实现优化和重构过程模型。该算法在四个典型测试函数上进行了对比实验,结果表明该算法在流过程挖掘中具有更好的收敛性和稳定性。...
Saved in:
| Published in | 计算机工程与科学 Vol. 39; no. 5; pp. 897 - 903 |
|---|---|
| Main Author | |
| Format | Journal Article |
| Language | Chinese |
| Published |
江苏开放大学信息与机电工程学院,江苏南京210017%南京航空航天大学计算机科学与技术学院,江苏南京,211106%江苏开放大学信息与机电工程学院,江苏南京,210017
2017
南京航空航天大学计算机科学与技术学院,江苏南京211106 |
| Subjects | |
| Online Access | Get full text |
| ISSN | 1007-130X |
| DOI | 10.3969/j.issn.1007-130X.2017.05.012 |
Cover
| Summary: | 过程流数据具有实时性、连续性和时序性等特征,使得传统过程挖掘算法难以发现隐含信息和演化过程。针对流过程模型的动态演化和重构要求,提出了一种基于时序行为分析的自适应混合启发式协同优化算法。首先定义演化流过程模型,基于日志活动间的隐含依赖关系改进过程逻辑的启发式挖掘规则,然后定义基于时序行为的老化因子,并引入高斯变异的多种群协作的自适应策略,改进粒子群优化算法的全局和局部精确寻优能力,实现优化和重构过程模型。该算法在四个典型测试函数上进行了对比实验,结果表明该算法在流过程挖掘中具有更好的收敛性和稳定性。 |
|---|---|
| Bibliography: | HUANG Li1,2 ,TAN Wen-an1 ,XU Xiao-yuan2 (1. School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 211106~ 2. School of Information and Eleetromechanical Engineering,Jiangsu Open University, Nanjing 210017,China) Stream data possesses real-time, continuous and sequential features. To detect implicit in- formation and dynamic process in stream data, we propose a hybrid heuristic cooperative optimization al- gorithm based on time series prediction with self-adapting for stream process reengineering. Firstly, we define the stream process model, and improve the heuristic miner rules in the process logic based on the implicit dependency relation among log activities. Secondly, we define the ageing factor based on sequen- tial behaviors and introduce the multiple particle swarm cooperation self-adapting strategy based on Gauss mutation to improve the local and global search capacity of the PSO algorithm, thus the process model is optimized and reengineered. Comparative |
| ISSN: | 1007-130X |
| DOI: | 10.3969/j.issn.1007-130X.2017.05.012 |