Automatically parsing sports movement data is a complex topic that requires extensive domain knowledge. Although smart sports watches and activity trackers are recording similar dimensions, the data is often pre-processed out of the box and tracks generated by different systems can rarely be directly compared.
It is necessary to develop a tool that allows interative adjustment of the processing workflow to account for differences in data sets
Develop a Java tool for processing tcx cycling (movement) data as a multivariate time series.
The application is based on a machine learning framework that can be used for development. The prototype should be delivered as a standalone application.
More information can be found on the respective open source repositories: