基于趋势的时间序列分段线性化算法
|
摘要
通过分析时间序列的几何形态特征,研究时间序列向上、向下趋势的几何形态。根据时间序列的变化特征,提出了高、低滤波点和高、低滤波线概念,利用高低滤波线判断时间序列的向上趋势、向下趋势,提出了一种基于时间序列变化趋势的分段线性化方法。实验结果表明,这样分段线性化便于实现,运行速度快,保持了时间序列的形态特征,有较好的逼近性及线段个数也很少等优点。
The geometric form of upward and downward trends in time series was studied by analyzing the geometric characteristics of time series.Concepts of high or low filtering points and high or low filtering lines were proposed according to the variation characteristics of time series.These concepts were used to judge the upward trends and the downward trends of the time series.Furthermore,apiecewise linear representation of time series based on upward or downward property was proposed.The results of experiments show that this method is easy to be programmed.It has desirable approximation,runs fast,and keeps the shape of time series.Furthermore,the number of lines is small.
The geometric form of upward and downward trends in time series was studied by analyzing the geometric characteristics of time series.Concepts of high or low filtering points and high or low filtering lines were proposed according to the variation characteristics of time series.These concepts were used to judge the upward trends and the downward trends of the time series.Furthermore,apiecewise linear representation of time series based on upward or downward property was proposed.The results of experiments show that this method is easy to be programmed.It has desirable approximation,runs fast,and keeps the shape of time series.Furthermore,the number of lines is small.
引文
[1]Bogusz J,Klos A,Bos M S,et al.On the impact of a quadratic acceleration term in the analysis of position time series[C]∥EGU General Assembly Conference.US:IEEE,2016.
[2]Mann M E.On smoothing potentially non-stationary climate time series[J].Geophysical Research Letters,2013,31(7):10-1029.
[3]Ziegel E R.Analysis of financial time series[J].Technometrics,2010,44(4):408-408.
[4]Jaeger H.Observable operator models for discrete stochastic time series[J].Neural Computation,2014,12(6):1371-1398.
[5]张晓琴,陈蜀宇.新颖的面向网络服务的动态信任模型[J].重庆大学学报,2013,36(4):56-63.ZHANG Xiaoqin,Chen S.A novel dynamic trust model for network service[J].Journal of Chongqing University,2013,36(4):56-63.(in Chinese)
[6]Tokinaga S,Moriyasu H,Miyazaki A,et al.Forecasting of time series with fractal geometry by using scale transformations and parameter estimations obtained by the wavelet transform[J].Electronics&Communications in Japan,2015,80(8):20-30.
[7]Fuentes J,Poncela P,Rodriguez J.Sparse partial least squares in time series for macroeconomic forecasting[J].Journal of Applied Econometrics,2015,30(4):576-595.
[8]Holst U,H9ssjer O,Bj9rklund C,et al.Locally weighted least squares kernel regression and statistical evaluation of lidar measurements[J].Environmetrics,2015,7(4):401-416.
[9]KEOGH E.A fast and robust method for pattern matching in time series databases[C]∥proceedings of the 9th International conference on Tools with Artificial Intelligence.Newport Beach:IEEE,1997:578-584.
[10]尚福华,孙达辰.基于时间序列趋势转折点的分段线性表示[J].计算机应用研究,2010,27(6):2075-2077.SHANG Fuhua,SUN Dachen.PLR based on time series tendency turning point[J].Application Research of Computers,2010,27(6):2075-2074.(in Chinese)
[11]周黔,吴铁军.基于重要点的时间序列趋势特征提取方法[J].浙江大学学报(工学版),2007,4(11):1782-1787.ZHOU Qian,WU Tiejun.Trend feature extraction method based on important points in time series[J].Journal of Zhejiang University,2007,41(11):1782-1787.(in Chinese)
[12]Li L,Su X,Zhang Y,et al.Trend modeling for traffic time series analysis:an integrated study[J].IEEE Transactions on Intelligent Transportation Systems,2015,16(6):3430-3439.
[13]林意,孔斌强.基于多尺度的时间序列固定分段数线性表示[J].计算机工程与应用,2016,52(21):81-87.LIN Yi,KONG Binqiang.Time series piecewise linear representation of fixed section number based on multi scale[J].Computer Engineering&Applications,2016,52(21):81-87.(in Chinese)
[14]林意,王智博.基于一阶滤波的时间序列分段线性表示方法[J].计算机工程,2016,42(9):151-157.LIN Yi,WANG Zhibo.Time series piecewise linear representation method based on first-order filtering[J].Computer Engineering,2016,42(9):151-157.(in Chinese)
[15]詹艳艳,徐荣聪,陈晓云.基于斜率提取边缘点的时间序列分段线性表示方法[J].计算机科学,2006,33(11):139-142.ZHAN Yanyan,XU Rongcong,CHEN Xiaoyun.Time series piecewise linear representation based on slope extract edge point[J].Computer Science,2006,33(11):139-142.(in Chinese)
[16]戴爱明,高学东.时间序列三角极值点线性分段算法[J].南昌航空大学学报(自然科学版),2009,23(2):25-28.DAI Aiming,GAO Xuedong.A linear segmentation algorithm for time series based on triangle extreme points[J].Journal of Nanchang Hangkong University,2009.(in Chinese)
[2]Mann M E.On smoothing potentially non-stationary climate time series[J].Geophysical Research Letters,2013,31(7):10-1029.
[3]Ziegel E R.Analysis of financial time series[J].Technometrics,2010,44(4):408-408.
[4]Jaeger H.Observable operator models for discrete stochastic time series[J].Neural Computation,2014,12(6):1371-1398.
[5]张晓琴,陈蜀宇.新颖的面向网络服务的动态信任模型[J].重庆大学学报,2013,36(4):56-63.ZHANG Xiaoqin,Chen S.A novel dynamic trust model for network service[J].Journal of Chongqing University,2013,36(4):56-63.(in Chinese)
[6]Tokinaga S,Moriyasu H,Miyazaki A,et al.Forecasting of time series with fractal geometry by using scale transformations and parameter estimations obtained by the wavelet transform[J].Electronics&Communications in Japan,2015,80(8):20-30.
[7]Fuentes J,Poncela P,Rodriguez J.Sparse partial least squares in time series for macroeconomic forecasting[J].Journal of Applied Econometrics,2015,30(4):576-595.
[8]Holst U,H9ssjer O,Bj9rklund C,et al.Locally weighted least squares kernel regression and statistical evaluation of lidar measurements[J].Environmetrics,2015,7(4):401-416.
[9]KEOGH E.A fast and robust method for pattern matching in time series databases[C]∥proceedings of the 9th International conference on Tools with Artificial Intelligence.Newport Beach:IEEE,1997:578-584.
[10]尚福华,孙达辰.基于时间序列趋势转折点的分段线性表示[J].计算机应用研究,2010,27(6):2075-2077.SHANG Fuhua,SUN Dachen.PLR based on time series tendency turning point[J].Application Research of Computers,2010,27(6):2075-2074.(in Chinese)
[11]周黔,吴铁军.基于重要点的时间序列趋势特征提取方法[J].浙江大学学报(工学版),2007,4(11):1782-1787.ZHOU Qian,WU Tiejun.Trend feature extraction method based on important points in time series[J].Journal of Zhejiang University,2007,41(11):1782-1787.(in Chinese)
[12]Li L,Su X,Zhang Y,et al.Trend modeling for traffic time series analysis:an integrated study[J].IEEE Transactions on Intelligent Transportation Systems,2015,16(6):3430-3439.
[13]林意,孔斌强.基于多尺度的时间序列固定分段数线性表示[J].计算机工程与应用,2016,52(21):81-87.LIN Yi,KONG Binqiang.Time series piecewise linear representation of fixed section number based on multi scale[J].Computer Engineering&Applications,2016,52(21):81-87.(in Chinese)
[14]林意,王智博.基于一阶滤波的时间序列分段线性表示方法[J].计算机工程,2016,42(9):151-157.LIN Yi,WANG Zhibo.Time series piecewise linear representation method based on first-order filtering[J].Computer Engineering,2016,42(9):151-157.(in Chinese)
[15]詹艳艳,徐荣聪,陈晓云.基于斜率提取边缘点的时间序列分段线性表示方法[J].计算机科学,2006,33(11):139-142.ZHAN Yanyan,XU Rongcong,CHEN Xiaoyun.Time series piecewise linear representation based on slope extract edge point[J].Computer Science,2006,33(11):139-142.(in Chinese)
[16]戴爱明,高学东.时间序列三角极值点线性分段算法[J].南昌航空大学学报(自然科学版),2009,23(2):25-28.DAI Aiming,GAO Xuedong.A linear segmentation algorithm for time series based on triangle extreme points[J].Journal of Nanchang Hangkong University,2009.(in Chinese)