Description
The DEIM-CUR matrix factorization has been demonstrated to be a viable subset selection tool in the electrocardiogram. The CUR factorization, however, can be formed in a variety of ways. We present a comparison of some of these CUR methods with some commonly used clustering algorithms, evaluating each method's performance on three different types of data. In doing so, we demonstrate the utility of CUR index selection schemes in data subset selection.
Learning Objective: After viewing this poster, the learner should be better able to 1) understand some of the pros and cons of using CUR index selection schemes for data subset selection, and 2) recognize that CUR index selection schemes are viable methods for subset selection from among different data types.
Authors:
Emily Hendryx (Presenter)
University of Central Oklahoma
Nabil Chaabane, Rice University
Beatrice Riviere, Rice University
Craig Rusin, Baylor College of Medicine
Presentation Materials: