多变量时间序列比较 [英] Multivariate Time series comparing

问题描述

大家好吧

我有四个功能和五个Sample这个功能的数字，并在不同的运行中为特定的应用程序提取。

假设我的数据跟随形状：

for first Run

Hi anyone
I have four feature and five Sample for this features that is numeric and extracted for specific application in different runs.
assumes that my data is following shape:
for first Run

	Feature			
	F1#	F2#	F3#	F4#
Sample1	15	142	500	310
Sample2	20	167	601	490
Sample3	25	189	699	670
Sample4	29	199	748	820
Sample5	55	256	799	1000

第二次运行

for second run

	Feature			
	F1#	F2#	F3#	F4#
Sample1	26	180	568	412
Sample2	34	192	672	499
Sample3	39	201	721	899
Sample4	49	222	801	1012
Sample5	77	291	900	2800

我的问题是：

如何比较这些数据？

机器学习方法或相似度测量方法适用于此多变量时间序列？

事实上，我想说这些数据是相似的，属于一个应用程序。

真诚的你。

My question is :
how compare this data?
what Machine learning method or Similarity measurement approach is appropriate for this Multivariate Time series?
In fact i want to say this data are similar and belongs for one applications.
Sincerely yours.

推荐答案

您可以在这里学习背景知识：

http://en.wikipedia.org/wiki/Multivariate_analysis_of_variance [ ^ ]，

http://en.wikipedia.org/wiki/ Multivariate_statistics [ ^ ]，

http://www.birs.ca/workshops/2006/06w5054/files/ byron-roe.pdf [ ^ ]，

http：// christophe。 genolini.free.f r / recherche / atelierATelecharger / Jose％20cortinas.pdf [ ^ ]，

http： //www.units.miamioh.edu/sumsri/sumj/2000/stat00.pdf [ ^ ]。



或许，你可以想找到一些可用的代码。要问问题的编程方面，标记Linux是不够的，并不是主要的事情;你还需要标记你想要使用的语言，或者其他一些细节。

-SA

You can learn the background, for example, here:
http://en.wikipedia.org/wiki/Multivariate_analysis_of_variance[^],
http://en.wikipedia.org/wiki/Multivariate_statistics[^],
http://www.birs.ca/workshops/2006/06w5054/files/byron-roe.pdf[^],
http://christophe.genolini.free.fr/recherche/atelierATelecharger/Jose%20cortinas.pdf[^],
http://www.units.miamioh.edu/sumsri/sumj/2000/stat00.pdf[^].

Probably, you can want to find some available code. To ask about programming aspects of the problem, tagging "Linux" is not enough and is not a main thing; you would also need to tag the language you want to use, perhaps some other detail.

—SA

这篇关于多变量时间序列比较的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！