Skip to content


Folders and files

Last commit message
Last commit date

Latest commit



6 Commits

Repository files navigation

Over-Time Stability Evaluation

ots-eval is a toolset for the over-time stability evaluation of multiple multivariate time series based on cluster transitions. It contains an over-time stability measure for crisp over-time clusterings called CLOSE [1], one stability measure for fuzzy over-time clusterings called FCSETS [2], two outlier detection algorithms DOOTS [3,4] and DACT [5] addressing cluster-transition-based outliers and an over-time clustering algorithm named C(OTS)^2 [6]. All approaches focus on multivariate time series data that is clustered per timestamp.

The toolset was implemented by Martha Krakowski (Tatusch) and Gerhard Klassen.


You can simply install ots-eval by using pip:

pip install ots-eval

You can import the package in your Python script via:

import ots_eval


ots-eval requires:

  • python>=3.7
  • pandas>=1.0.0
  • numpy>=1.19.2
  • scipy>=1.3.0


In the doc folder, there are some explanations for the usage of every approach.


In the data folder, the generated data sets from the listed publications are provided. A short explanation can be found in the doc folder:


ots-eval is distributed under the 3-Clause BSD license.


This toolset is the implementation of approaches from our following works:

[1] Tatusch, M., Klassen, G., Bravidor, M., and Conrad, S. (2020).
How is Your Team Spirit? Cluster Over-Time Stability Evaluation.
In: Machine Learning and Data Mining in Pattern Recognition, 16th International Conference on Machine Learning and Data Mining, MLDM 2020, pages 155–170.

[2] Klassen, G., Tatusch, M., Himmelspach, L., and Conrad, S. (2020).
Fuzzy Clustering Stability Evaluation of Time Series.
In: Information Processing and Management of Uncertainty in Knowledge-Based Systems, 18th International Conference, IPMU 2020, pages 680-692.

[3] Tatusch, M., Klassen, G., Bravidor, M., and Conrad, S. (2019).
Show me your friends and i’ll tell you who you are. Finding anomalous time series by conspicuous clus- ter transitions.
In: Data Mining. AusDM 2019. Communications in Computer and Information Science, pages 91–103.

[4] Tatusch, M., Klassen, G., and Conrad, S. (2020).
Behave or be detected! Identifying outlier sequences by their group cohesion.
In: Big Data Analytics and KnowledgeDiscovery, 22nd International Conference, DaWaK 2020, pages 333–347.

[5] Tatusch, M., Klassen, G., and Conrad, S. (2020).
Loners stand out. Identification of anomalous subsequences based on group performance.
In: Advanced Data Mining and Applications, ADMA 2020, pages 360–369.

[6] Klassen, G., Tatusch, M., and Conrad, S. (2020).
Clustering of time series regarding their over-time stability.
In: Proceedings of the 2020 IEEE Symposium Series on Computational Intelligence (SSCI), pages 1051–1058.

[7] Klassen, G., Tatusch, M., and Conrad, S. (2021).
Cluster-Based Stability Evaluation in Time Series Data Sets.
In: (submitted).


No description, website, or topics provided.







No packages published