With the improvement of measurement devices, the capturing and subsequent analysis of multiple physiologic measures has increased significantly. One such area is the scientific research of reflex response of the human nervous system. However, these responses are dependent on multiple other factors, such as body position.
Physiologic measurements are described as a multivariate time series, meaning that for each available time step, multiple measurements are performed. The main problem is now to find correlations and relationships among those measurements. The computation of simple correlations in a pair-wise fashion often remains without result. These visual techniques are required to help domain experts to gain insight into their data.
- Design and implement a visual analytics system for multivariate time series
- Correlation of time series using statistical, interactive, and visual methods
- Similarity-based exploration using Focus+Context and detail on-demand techniques
- Visual aggregation of time series for different groups (e.g., trials) and clusters
- High interest in the topic
- Knowledge about Time Series Visualization, Statistics, and Machine Learning
- Excellent programming skills: Java or Python, React or Angular, D3.js
- Scope: Bachelor or Master
- 3 Month Project, 6 Month Thesis
- Start: immediately
- Jürgen Bernard, Nils Wilhelm, Maximilian Scherer, Thorsten May, Tobias Schreck. TimeSeriesPaths: Projection-Based Explorative Analysis of Multivarate Time Series Data. J. WSCG Vol. 20, No. 2, 97-106, 2012. URL: http://wscg.zcu.cz/jwscg/J\_WSCG\_2012/!\_2012-Journal-Full-2.pdf
- Ming C. Hao, Manish Marwah, Halldòr Janetzko, Ratnesh Sharma, Daniel A. Keim, Umeshwar Dayal, Debprakash Patnaik, Naren Ramakrishnan. Visualizing frequent patterns in large multivariate time series. Visualization and Data Analysis, SPIE Proceedings, Vol. 7868, SPIE, 2011. URL: doi.org/10.1117/12.872169
- Flor de Luz Palomino Valdivia, Herwin Alayn Huillcen Baca, Ana María Cuadros Valdivia. Guided Visual Analysis of Multivariate Time Series. Advances in Information and Communication, Springer International Publishing, 247-262, 2022. URL: link.springer.com/chapter/10.1007/978-3-030-98012-2_19