Time-series Visualization with the Matrix Profile and Multidimensional Scaling.
Project description
TimeSeriesVisualization
Visualization of Time-Series with the Matrix Profile and Multidimensional Scaling - a Python Implementation.
Introduction
Multidimensional scaling is an algorithm that projects a set of objects represented by their distance matrix into a lesser dimensional space, such that the pairwise distance is preserved as much as possible. If we project onto a 1-dimensional or a 2-dimensional space, we can even plot the resultant projections, and visually understand the similarity between these objects.
At a first glance, this seems applicable to time-series as well, where we want to capture similar regions across the entire time-series.
However, since we do not know apriori where our regions of interest are, we can simply select all subsequences of a given length from the original time-series.
Let's try this out!
References
Matrix Profile III: The Matrix Profile Allows Visualization of Salient Subsequences in Massive Time Series.
Chin-Chia Michael Yeh, Helga Van Herle, and Eamonn Keogh. IEEE ICDM 2016.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for tsvisualize-1.0-py2-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | b9d23f805b713d05b0282898ca6f55b3dd155e0a47f9ce6816422544563c8184 |
|
MD5 | 59a81508dfbec8a26697328178f5e9ec |
|
BLAKE2b-256 | 3d74edff0062f282bbb5ce0bd5185666f611fc23f9a88ef7c5ee83977c8a1a3a |