模型聚合解聚的智能触发机制
|
摘要
针对现有模型聚合解聚(AD)触发机制人工依赖性高、频繁聚合解聚的问题,提出了一种基于关注域的多实体时序离群点检测算法的智能触发机制。首先,基于关注近邻划分关注域;然后,计算关注域中实体的k距离离群值,得到关注域的离群值;最后,结合一种基于最强关注域阈值判定方法,构建聚合解聚触发机制。在真实数据集上的实验结果表明,与传统的单实体时序离群点检测算法相比,所提算法在指标Precision、Recall和综合指标F1-score上均提升了10个百分点以上,不仅能及时地判断聚合解聚操作的触发时机,而且能使得仿真系统智能地检测出发生突发情况的仿真实体,满足了多分辨率建模的要求。
Aiming at high manual dependence and frequent Aggregation and Disaggregation(AD) of existing model AD trigger mechanisms, an intelligent trigger mechanism based on focus-area multi-entity temporal outlier detection algorithm was proposed. Firstly, the focus-areas were divided based on attention neighbors. Secondly, the outlier score of focus-area was obtained by calculating the k-distance outlier score of entities in a focus-area. Finally, a trigger mechanism for AD was constructed based on strongest-focus-area threshold decision method. The experimental results on real dataset show that, compared with the traditional single-entity temporal outlier detection algorithms, the proposed algorithm improves the performance of Precision, Recall and F1-score by more than 10 percentage points. The proposed algorithm can not only judge the trigger time of the AD operation in time, but also enable the simulation system to intelligently detect the simulation entities with emergency situation and meet the requirements of multi-resolution modeling.
Aiming at high manual dependence and frequent Aggregation and Disaggregation(AD) of existing model AD trigger mechanisms, an intelligent trigger mechanism based on focus-area multi-entity temporal outlier detection algorithm was proposed. Firstly, the focus-areas were divided based on attention neighbors. Secondly, the outlier score of focus-area was obtained by calculating the k-distance outlier score of entities in a focus-area. Finally, a trigger mechanism for AD was constructed based on strongest-focus-area threshold decision method. The experimental results on real dataset show that, compared with the traditional single-entity temporal outlier detection algorithms, the proposed algorithm improves the performance of Precision, Recall and F1-score by more than 10 percentage points. The proposed algorithm can not only judge the trigger time of the AD operation in time, but also enable the simulation system to intelligently detect the simulation entities with emergency situation and meet the requirements of multi-resolution modeling.
引文
[1]李明忠,毕长剑,刘小荷,等.空军作战仿真模型聚合与解聚研究[J].系统仿真学报,2008,20(14):3679-3684.(LI M Z,BIC J,LIU X H,et al.Research on model aggregation and disaggregation for air force combat simulation[J].Journal of System Simulation,2008,20(14):3679-3684.)
[2]李凤霞,卢兆涵,雷正朝,等.基于队形的聚合解聚方法研究[J].系统仿真学报,2013,25(10):2308-2313.(LI F X,LU ZH,LEI Z Z,et al.Aggregation and disaggregation methods research based on formation[J].Journal of System Simulation,2013,25(10):2308-2313.)
[3]GOU C X,CAI B G.Multi-resolution entities aggregation and disaggregation method for train control system modeling and simulation based on HLA[C]//ITSC 2014:Proceedings of the 2014 17th International IEEE Conference on Intelligent Transportation Systems.Piscataway,NJ:IEEE,2014:2367-2372.
[4]朱敏洁,周深根.多分辨率模型转换的触发机制[J].指挥控制与仿真,2015,37(5):44-47.(ZHU M J,ZHOU S G.Multi-resolution combat model triggering mechanism[J].Command Control&Simulation,2015,37(5):44-47.)
[5]AGGARWAL C C.Outlier Analysis[M].2 nd ed.Berlin:Springer,2016:286.
[6]KATHAREIOS G,ANGHEL A,MATE A,et al.Catch it if you can:real-time network anomaly detection with low false alarm rates[C]//ICMLA 2017:Proceedings of the 16th IEEE International Conference on Machine Learning and Applications.Piscataway,NJ:IEEE,2017:924-929.
[7]BENKABOU S E,BENABDESLEM K,CANITIA B.Unsupervised outlier detection for time series by entropy and dynamic time warping[J].Knowledge and Information Systems,2018,54(2):463-486.
[8]YANG C L,LIAO W J.Adjacent Mean Difference(AMD)method for dynamic segmentation in time series anomaly detection[C]//SII2017:Proceedings of the 2017 IEEE/SICE International Symposium on System Integration.Piscataway,NJ:IEEE,2017:241-246.
[9]KHA N H,ANH D T.From cluster-based outlier detection to time series discord discovery[M]//LI X L,CAO T,LIM E P,et al.Trends and Applications in Knowledge Discovery and Data Mining,LNCS 9441.Cham:Springer,2015:16-28.
[10]REN H R,LIU M M,LI Z W,et al.A piecewise aggregate pattern representation approach for anomaly detection in time series[J].Knowledge-Based Systems,2017,135:29-39.
[11]ZHENG D Q,LI F H,ZHAO T J.Self-adaptive statistical process control for anomaly detection in time series[J].Expert Systems with Applications,2016,57:324-336.
[12]ZENG J,ZHANG L,SHI G T,et al.An ARIMA based real-time monitoring and warning algorithm for the anomaly detection[C]//ICPADS 2017:Proceedings of the 2017 IEEE 23rd International Conference on Parallel and Distributed Systems.Piscataway,NJ:IEEE,2017:469-476.
[13]NA G S,KIM D H,YU H.DILOF:effective and memory efficient local outlier detection in data streams[C]//KDD 2018:Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery&Data Mining.New York:ACM,2018:1993-2002.
[14]WANG L,XU L Y,XUE Y L,et al.Group behavior time series anomaly detection in specific network space based on separation degree[J].Cluster Computing,2016,19(3):1201-1210.
[15]AHMAD S,LAVIN A,PURDY S,et al.Unsupervised real-time anomaly detection for streaming data[J].Neurocomputing,2017,262:134-147.
[16]HAN J,KAMBER M,PEI J.Data Mining:Concepts and Techniques[M].3rd ed.San Francisco,CA:Morgan Kaufmann Publishers,2011:351-376.
[17]LPL职业联赛.2018LPL春季赛季后赛RNG vs WE第二局[EB/OL].[2018-09-12].https://v.qq.com/x/cover/191162cgjvuzbxm/p00261fyo6v.html.(League of Legends Pro League.Royal Never Give Up vs.Team WE:LPL 2018 Spring Playoffs-Round 2[EB/OL].[2018-09-12].https://v.qq.com/x/cover/191162cgjvuzbxm/p00261fyo6v.html.)
[18]SATHE S,AGGARWAL C C.Subspace outlier detection in linear time with randomized hashing[C]//ICDM 2016:Proceedings of the 2016 IEEE 16th International Conference on Data Mining.Piscataway,NJ:IEEE,2016:459-468.
[19]MARTINS H,PALMA L,CARDOSO A,et al.A support vector machine based technique for online detection of outliers in transient time series[C]//ASCC 2015:Proceedings of the 10th Asian Control Conference.Piscataway,NJ:IEEE,2015:1-6.
[20]NING J,CHEN L T,ZHOU C,et al.Parameter k search strategy in outlier detection[J].Pattern Recognition Letters,2018,112:56-62.
[2]李凤霞,卢兆涵,雷正朝,等.基于队形的聚合解聚方法研究[J].系统仿真学报,2013,25(10):2308-2313.(LI F X,LU ZH,LEI Z Z,et al.Aggregation and disaggregation methods research based on formation[J].Journal of System Simulation,2013,25(10):2308-2313.)
[3]GOU C X,CAI B G.Multi-resolution entities aggregation and disaggregation method for train control system modeling and simulation based on HLA[C]//ITSC 2014:Proceedings of the 2014 17th International IEEE Conference on Intelligent Transportation Systems.Piscataway,NJ:IEEE,2014:2367-2372.
[4]朱敏洁,周深根.多分辨率模型转换的触发机制[J].指挥控制与仿真,2015,37(5):44-47.(ZHU M J,ZHOU S G.Multi-resolution combat model triggering mechanism[J].Command Control&Simulation,2015,37(5):44-47.)
[5]AGGARWAL C C.Outlier Analysis[M].2 nd ed.Berlin:Springer,2016:286.
[6]KATHAREIOS G,ANGHEL A,MATE A,et al.Catch it if you can:real-time network anomaly detection with low false alarm rates[C]//ICMLA 2017:Proceedings of the 16th IEEE International Conference on Machine Learning and Applications.Piscataway,NJ:IEEE,2017:924-929.
[7]BENKABOU S E,BENABDESLEM K,CANITIA B.Unsupervised outlier detection for time series by entropy and dynamic time warping[J].Knowledge and Information Systems,2018,54(2):463-486.
[8]YANG C L,LIAO W J.Adjacent Mean Difference(AMD)method for dynamic segmentation in time series anomaly detection[C]//SII2017:Proceedings of the 2017 IEEE/SICE International Symposium on System Integration.Piscataway,NJ:IEEE,2017:241-246.
[9]KHA N H,ANH D T.From cluster-based outlier detection to time series discord discovery[M]//LI X L,CAO T,LIM E P,et al.Trends and Applications in Knowledge Discovery and Data Mining,LNCS 9441.Cham:Springer,2015:16-28.
[10]REN H R,LIU M M,LI Z W,et al.A piecewise aggregate pattern representation approach for anomaly detection in time series[J].Knowledge-Based Systems,2017,135:29-39.
[11]ZHENG D Q,LI F H,ZHAO T J.Self-adaptive statistical process control for anomaly detection in time series[J].Expert Systems with Applications,2016,57:324-336.
[12]ZENG J,ZHANG L,SHI G T,et al.An ARIMA based real-time monitoring and warning algorithm for the anomaly detection[C]//ICPADS 2017:Proceedings of the 2017 IEEE 23rd International Conference on Parallel and Distributed Systems.Piscataway,NJ:IEEE,2017:469-476.
[13]NA G S,KIM D H,YU H.DILOF:effective and memory efficient local outlier detection in data streams[C]//KDD 2018:Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery&Data Mining.New York:ACM,2018:1993-2002.
[14]WANG L,XU L Y,XUE Y L,et al.Group behavior time series anomaly detection in specific network space based on separation degree[J].Cluster Computing,2016,19(3):1201-1210.
[15]AHMAD S,LAVIN A,PURDY S,et al.Unsupervised real-time anomaly detection for streaming data[J].Neurocomputing,2017,262:134-147.
[16]HAN J,KAMBER M,PEI J.Data Mining:Concepts and Techniques[M].3rd ed.San Francisco,CA:Morgan Kaufmann Publishers,2011:351-376.
[17]LPL职业联赛.2018LPL春季赛季后赛RNG vs WE第二局[EB/OL].[2018-09-12].https://v.qq.com/x/cover/191162cgjvuzbxm/p00261fyo6v.html.(League of Legends Pro League.Royal Never Give Up vs.Team WE:LPL 2018 Spring Playoffs-Round 2[EB/OL].[2018-09-12].https://v.qq.com/x/cover/191162cgjvuzbxm/p00261fyo6v.html.)
[18]SATHE S,AGGARWAL C C.Subspace outlier detection in linear time with randomized hashing[C]//ICDM 2016:Proceedings of the 2016 IEEE 16th International Conference on Data Mining.Piscataway,NJ:IEEE,2016:459-468.
[19]MARTINS H,PALMA L,CARDOSO A,et al.A support vector machine based technique for online detection of outliers in transient time series[C]//ASCC 2015:Proceedings of the 10th Asian Control Conference.Piscataway,NJ:IEEE,2015:1-6.
[20]NING J,CHEN L T,ZHOU C,et al.Parameter k search strategy in outlier detection[J].Pattern Recognition Letters,2018,112:56-62.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |