This toolbox provides Matlab implementations of Continuous Profile Models
(CPM) for alignment and normalization of time series data. Specifically,
two models are provided (well, one so far), the EM-CPM, from the paper
"Multiple Alignment of Continuous Time Series", and the HB-CPM, from the
paper "Bayesian Detection of Infrequent Differences in Sets of Time Series
with Shared Structure". The HB-CPM implementation is not yet available.
Both of these models are reported in my PhD thesis, which is listed along
with other references, below.
Conditions of Use:
This toolbox is available free for educational and research use. All use of
this software is at the user's own risk.
Please cite my thesis if using this toolbox.
Before downloading the toolbox, you may want to read the
README file, which contains some information about using the CPM
Matlab toolbox. You will also need to read relevant portions of the literature
to understand the toolbox.
If after reading the README above, you want to download the code package,
(which includes Matlab code and demo scripts). The data which goes with the demo scripts
can be found separately (because of the large size),
here. There is a data
README file in that directory explaining the various data sets.
Analysis of sibling time series data: alignment and difference
(abstract and thesis)
Ph.D. Thesis, Department of Computer Science, University of Toronto, 2006.
Multiple Alignment of Continuous Time Series.
(abstract, paper, slides, and audio demo)
Jennifer Listgarten, Radford M. Neal, Sam T. Roweis and Andrew Emili,
in Advances in Neural Information Processing Systems 17,
MIT Press, Cambridge, MA, 2005 (NIPS 2004).
Bayesian Detection of Infrequent Differences in
Sets of Time Series with Shared Structure.
Jennifer Listgarten, Radford M. Neal, Sam T. Roweis, Rachel Puckrin and Sean Cutler,
To appear in Advances in Neural Information Processing Systems 19, MIT Press,
Cambridge, MA, 2007 (NIPS 2006).
Best Student Paper, Honorable Mention
October 12th, 2006 -- First version released. This contains an
implementation of the EM-CPM, along with scripts which show how to use
it, and some demo data that can be used with the scripts. The HB-CPM is not