一种基于多阈值模板的快速分类在线检测方法

doi:10.16180/j.cnki.issn1007-7820.2024.06.010

电子科技 ›› 2024, Vol. 37 ›› Issue (6): 77-83.doi: 10.16180/j.cnki.issn1007-7820.2024.06.010

一种基于多阈值模板的快速分类在线检测方法

薛宇鑫, 齐金鹏, 贾灿, 袁傲, 黄莉娜

东华大学信息科学与技术学院,上海 201620

收稿日期:2023-01-15 出版日期:2024-06-15 发布日期:2024-06-20
作者简介:薛宇鑫(1994-),男,硕士研究生。研究方向:大数据异常检测。
齐金鹏(1977-),男,博士,副教授。研究方向:大数据异常检测、图像处理。
基金资助:
国家自然科学基金(61305081);国家自然科学基金(61104154);上海市自然科学基金(16ZR1401300);上海市自然科学基金(16ZR1401200)

A Fast Classification Online Detection Method Based on Multi-Threshold Template

XUE Yuxin, QI Jinpeng, JIA Can, YUAN Ao, HUANG Lina

College of Information Science & Technology,Donghua University,Shanghai 201620,China

Received:2023-01-15 Online:2024-06-15 Published:2024-06-20
Supported by:
National Natural Science Foundation of China(61305081);National Natural Science Foundation of China(61104154);Shanghai Natural Science Foundation of China(16ZR1401300);Shanghai Natural Science Foundation of China(16ZR1401200)

摘要/Abstract

摘要：

传统离线数据分析方法对于处理即时性高和流量大的数据存在缺陷,而在线检测模型可以满足数据流分析的实时性要求。文中提出了一种基于多阈值模板的在线检测方法。该方法结合多路搜索树突变点检测(Ternary Search Tree and Kolmogorov-Smirnov,TSTKS)算法进行在线检测,基于突变点密度更新窗口长度从而提高了突变点检测精度。采用等量分级策略实现对时序数据的自学习、匹配和分类,进而对大规模病变数据进行状态检测和预测。仿真实验和病变数据的实验结果表明,所提方法具有效果高、分类准确等优点,为大规模时序数据进行快速分类研究提供了新方法。

关键词: 时序数据, TSTKS算法, 滑动窗口, 在线检测理论, 缓冲区, 突变点密度, 多阈值模板, 等量分级策略

Abstract:

The traditional off-line data analysis method has many shortcomings in processing the data with high immediacy and large flow, while the online detection model can meet the real-time requirements of data flow analysis. This study proposes an online detection method based on the multi-threshold template. The proposed method combines TSTKS(Ternary Search Tree and Kolmogorov-Smirnov) algorithm for online detection, and updates the window length based on the mutation point density to improve the mutation point detection accuracy. Self-learning, matching and classification of time series data are realized by equal grading strategy, so as to detect and predict the status of large-scale lesion data. The experimental results of simulation experiment and lesion data show that the proposed method has the advantages of high efficiency and accurate classification, which provides a new method for the rapid classification of large-scale time series data.

Key words: time series data, TSTKS algorithm, sliding window, online detection theory, buffer, mutation point density, multi-threshold template, equal grading police

中图分类号:

TP311

薛宇鑫, 齐金鹏, 贾灿, 袁傲, 黄莉娜. 一种基于多阈值模板的快速分类在线检测方法[J]. 电子科技, 2024, 37(6): 77-83.

XUE Yuxin, QI Jinpeng, JIA Can, YUAN Ao, HUANG Lina. A Fast Classification Online Detection Method Based on Multi-Threshold Template[J]. Electronic Science and Technology, 2024, 37(6): 77-83.

图/表 8

图1

图2

图3

图4

表1

表2

图5

图6

参考文献 20

[1]	刘晓艳, 王珏, 姚铁锤, 等. 基于时序数据处理的分布式光伏功率预测系统[J]. 数据与计算发展前沿, 2021, 3(4):140-148.
	Liu Xiaoyan, Wang Jue, Yao Tiechui, et al. A distributed photovoltaic power prediction system based on time series data processing[J]. Frontiers of Data and Computing, 2021, 3(4):140-148.
[2]	张长青, 杨楠. 基于车联网大数据分析的实时路况检测系统[J]. 电子科技, 2019, 32(8):66-69,74.
	Zhang Changqing, Yang Nan. Design of real-time road condition detection system based on big data analysisof vehicle network[J]. Electronic Science and Technology, 2019, 32(8):66-69,74.
[3]	艾辣椒. 基于多路搜索的时序突变异常快速分析[D]. 上海: 东华大学,2018:66-69.
	Ai Lajiao. A fast analysis for abrupt change detection on time series based on multipath search[D]. Shanghai: Donghua University,2018:66-69.
[4]	胡宇鹏. 时间序列数据挖掘中的特征表示与分类方法的研究[D]. 济南: 山东大学,2018:83-88.
	Hu Yupeng. Research on feature representation and classification methods in time series data mining[D]. Jinan: Shandong University,2018:83-88.
[5]	Casson A, Yates D, Smith S, et al. Wearable electroence-phalography[J]. Engineering in Medicine and Biology Magazine IEEE, 2010, 29(3):44-56.
[6]	Aljawarneh S, Anguera A, Atwood J W, et al. Particulari-ties of data mining in medicine:Lessons learned from patient medical time series data analysis[J]. Journal on Wireless Communications and Networking, 2019(1):1-29.
[7]	朱俊俊. 基于自适应随机策略的多突变点在线检测与病变数据关联分析[D]. 上海: 东华大学,2022:90-94.
	Zhu Junjun. Online detection of multiple change points and association analysis of pathological data based on adaptive random strategy[D]. Shanghai: Donghua University,2022:90-94.
[8]	邹俊晨. 一种基于自适应滑动窗口的突变点在线检测方法[D]. 上海: 东华大学,2019:69-73.
	Zou Junchen. An online change point detection method based on adaptive sliding window[D]. Shanghai: Donghua University,2019:69-73.
[9]	孔凡书, 齐金鹏, 龚汉鑫, 等. 一种基于波动向量分级技术的病变数据快速分析方法[J]. 电子科技, 2022, 35(7):1-6.
	Kong Fanshu, Qi Jinpeng, Gong Hanxin, et al. A fast analysis method of pathological data based on wave vector grading technology[J]. Electronic Science and Technology, 2022, 35(7):1-6.
[10]	龚汉鑫, 齐金鹏, 孔凡书, 等. 一种基于模板匹配与隶属度解析的时序异常快速检测方法[J]. 电子科技, 2022, 35(6):1-5,27.
	Gong Hanxin, Qi Jinpeng, Kong Fanshu, et al. A fast time series anomaly detection method based on template matching and membership analysis[J]. Electronic Science and Technology, 2022, 35(6):1-5,27.
[11]	孔凡书. 基于多突变点检测的波动向量分级技术研究与应用[D]. 上海: 东华大学,2021:53-60.
	Kong Fanshu. Research and application of a wave vector classification technique based on multiple change points detection[D]. Shanghai: Donghua University,2021:53-60.
[12]	Boukerche A, Zhang L, Alfandi O. Outlier detection:Methods,models,and classification[J]. ACM Computing Surveys, 2020, 53(3):1-37.
[13]	Bolgov M V. Bayesian estimation of the change point in the sequences of correlated random variables of hydrometeorological characteristic[J]. Russian Meteorology and Hydrology, 2021, 46(10):667-673.
[14]	展鹏. 基于时间序列挖掘的异常检测关键技术研究[D]. 济南: 山东大学,2020:88-94.
	Zhan Peng. Research on key technologies of anomaly detection based on time series mining[D]. Jinan: Shandong University,2020:88-94.
[15]	张友浩, 赵鸣, 徐梦瑶. 时序数据挖掘的预处理研究综述[J]. 智能计算机与应用, 2021, 11(1):74-78.
	Zhang Youhao, Zhao Ming, Xu Mengyao, et al. Summary of research on preprocessing of time series data mining[J]. Intelligent Computer and Applications, 2021, 11(1):74-78.
[16]	Qi J P, Zhang Q, Qi J, et al. Multi-channel detection for abrupt change based on the ternary search tree and kolmogorov statistic method[C]. Hangzhou: The Thirty- fourth Chinese Control Conference,2015:4968-4973.
[17]	Pearson R K. Outliers in process modeling and identification[J]. IEEE Transactions on Control Systems Technology, 2002, 10(1):55-63.
[18]	邬贺铨. 大数据时代的发展趋势[J]. 广东科技, 2016, 25(17):30-33.
	Wu Hequan. Development trend in the era of big data[J]. Guangdong Science and Technology, 2016, 25(17):30-33.
[19]	Alarcon-aquino V, Barria J A. Anomaly detection in communication networks using wavelets[J]. IEEE Proceedings Communications, 2002, 148(6):55-62.
[20]	贾凡, 严妍, 张家琪. 基于K-means聚类特征消减的网络异常检测[J]. 清华大学学报(自然科学版), 2018, 58(2):137-142.
	Jia Fan, Yan Yan, Zhang Jiaqi. K-means based feature reduction for network anomaly detection[J]. Journal of Tsinghua University(Science and Technology), 2018, 58(2):137-142.

策略	时耗/s	命中率/%	平均相对误差率/%	平均绝对误差率/%
a	0.218 4	0.673 3	0.070 1	0.104 1
b	0.121 0	0.980 0	0.105 6	0.107 8
c	0.131 6	0.995 0	0.101 2	0.101 7
d	0.196 5	0.998 3	0.050 8	0.050 9

初始缓冲区长度	初始滑动窗口长度	时耗/s	命中率/%	平均相对误差率/%	平均绝对误差率/%
2¹¹	2¹³	0.196 5	0.998 3	0.050 8	0.050 9
2¹²	2¹³	0.141 5	0.656 7	0.045 0	0.068 6
2¹¹	2¹⁴	0.162 7	0.915 0	0.066 8	0.073 0
2¹²	2¹⁴	0.131 7	0.643 3	0.035 1	0.054 6

一种基于多阈值模板的快速分类在线检测方法

A Fast Classification Online Detection Method Based on Multi-Threshold Template

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 20

相关文章 10

Metrics

本文评价

推荐阅读 0

[1]	贾灿, 齐金鹏, 袁傲, 薛宇鑫, 戴理. 基于多突变点与模板匹配的用电设备在线状态监测方法[J]. 电子科技, 2024, 37(6): 69-76.
[2]	袁傲, 齐金鹏, 贾灿, 薛宇鑫, 郭阳阳. 一种基于深度学习的时序病变数据段分类方法[J]. 电子科技, 2024, 37(6): 84-91.
[3]	孔凡书,齐金鹏,龚汉鑫,朱俊俊,曹一彤. 一种基于波动向量分级技术的病变数据快速分析方法[J]. 电子科技, 2022, 35(7): 1-6.
[4]	龚汉鑫,齐金鹏,孔凡书,朱俊俊,曹一彤. 一种基于模板匹配与隶属度解析的时序异常快速检测方法[J]. 电子科技, 2022, 35(6): 1-5.
[5]	朱俊俊,齐金鹏,钟金美,任晴,曹一彤. 一种基于随机交叠策略的多突变点在线检测方法[J]. 电子科技, 2022, 35(12): 1-9.
[6]	邹俊晨,齐金鹏,李娜,刘佳伦,朱厚杰. 一种快速的突变点在线检测算法设计与实现[J]. 电子科技, 2020, 33(8): 10-16.
[7]	侯严明,李菲菲,陈虬. 基于多特征融合的视频检索算法[J]. 电子科技, 2019, 32(5): 44-49.
[8]	张天奇,张顺康. 基于主成分分析的实时全网络异常检测方法[J]. 电子科技, 2019, 32(12): 17-21.
[9]	褚天,潘金满,杜磊. 基于自适应滑动窗口的RFID漏读数据清洗算法[J]. , 2016, 29(4): 24-.
[10]	江楠,徐秦. 数据流聚类算法在数据处理中的应用[J]. , 2015, 28(1): 155-.