This is a supporting page of the paper
Time Series Join on Correlation
Abdullah Mueen, Hossein Hamooni and Trilce Estrada
The paper is Here and the presentation slides are Here.
Joining two time series at their most correlated segment can provide important information about their synchrony. Finding the most correlated join segments is time consuming. In this work we have shown an algorithm that finds the best join segment for two long time series or two sets of time series efficiently at an interactive speed. Below we show a join for two currency conversion rates.
Code and Executables
We have a C++ code and a MATLAB function for the joining.
1. C++ code: There is an external dependency on fftw3.h which can be downloaded from here.
2. Matlab Function.
Spreadsheet of Experimental Results
We have compiled results of all the experiments in a spreadsheet. All time measurements are in seconds.
For scalability experiments,
RW data is generated by the following two lines in matlab : R = randn(1,10000); RW = cumsum(R);
Power Consumption
The dataset is taken from here.
JOCOR Based Clustering
The dataset is taken from here.
Oceanographic Time Series
The dataset is taken from here.
This page is created by:
Abdullah Mueen
Department of Computer Science, University of New Mexico