Supervised Time Series Pattern Discovery through Local Importance

This is a supporting page to our paper -  Supervised Time Series Pattern Discovery through Local Importance (TSPD)  

by Mustafa Gokce Baydogan and George Runger
*this study is presented in INFORMS 2012@Phoenix.  The presentation is available here. Note that there might be changes on the approach (as well as results) compared to our submission."
*the paper submitted to Knowledge and Information Systems will be available in Papers category in Files section."
 
DATASETS
We test our proposed approach on 45 datasets from UCR time series database  and 11 datasets from Hills et al.. The dataset characteristics for UCR datasets are in LPS supporting page. This table provides the information about the dataset characteristics (number of classes, length of time series, number of training and test instances). 
 
CODES
We implemented TSPD as functions from the LPS package in R. The source files for the package are provided here in the files section. You need to install R first. Package is compiled and experimented on 64bit Linux environment (Ubuntu 13.04). Recently it was tested on 64bit Windows 8 system (Oct 14, 2013). 
 
Install it from the local file by   'R CMD INSTALL yourfoldergoeshere/LPS_1.01.tar.gz'
 
After installing LPS package, R code here should be run.
 
HOW TO RUN TSPD
The details are described in the blog entry here.

Copyright © 2014 mustafa gokce baydogan

LinkedIn
Twitter
last.fm