PyTSML 1.1

Python implementations for a few machine learning algorithms used for working with time series, including LDMLT and DDE

PyTSML

Python Time Series Machine Learning

This package, which was initially made as a result of my MSC Thesis in Computer Engineering, contains Python implementations of few machine learning algorithms designed to work (mainly classify) data in the time series format.

Currently implemented methods:

• LDMLT (LogDet Divergence-Based Metric Learning With Triplet Constraints) [Mei, J., Liu, M., Karimi, H.R., & Gao, H. (2014). LogDet Divergence-Based Metric Learning With Triplet Constraints and Its Applications. IEEE Transactions on Image Processing, 23, 4920-4931.];
• DDE (Derivative Delay Embedding) [Zhang, Z., Song, Y., Wang, W., & Qi, H. (2016). Derivative Delay Embedding: Online Modeling of Streaming Time Series. Proceedings of the 25th ACM International on Conference on Information and Knowledge Management.];
• KNN classifier with DTW distance.

Usage

To use the package, build the wheel yourself, install it through pip or just use the source file in Your project.

pip3 install PyTSML

Contact

Please use Github issues page for anything related to this package.