Extra Colloquium Computer Science, Prof. B. Hammer (University of Bielefeld)

Date:                          Thursday, October 10^th 2013
Speaker:                     Prof. Barbara Hammer CITEC, University of Bielefeld, Germany
Room:                         5161.0267 (Bernoulliborg)
Time:                           16.00

Title: Discriminative dimensionality reduction and its applications

Abstract:

The amount of electronic data available today increases rapidly, such that humans rely on automated tools which allow them to intuitively scan data volumes for valuable information.

Dimensionality reducing data visualization, which displays high dimensional data in two or three dimensions, constitutes a popular tool to directly visualize data sets on the computer screen.

Dimensionality reduction is an inherently ill-posed problem, and the result of a dimensionality reduction tool largely varies depending on the chosen technology, the parameters, and partially even random aspects for non-deterministic algorithms.

Often, the reliability and suitability of the obtained visualization for the task at hand is not clear at all since a dimensionality reduction tool might focus on irrelevant aspects or noise in the data.

In the presentation, we discuss techniques how to enhance dimensionality reduction techniques by auxiliary information in the form of class labeling of the data, in particular proposing Fisher t-SNE as an extension of t-SNE to incorporate class information. This way, the visualization can concentrate on the aspects relevant for the given auxiliary information specified by the user rather than potential noise. We demonstrate the suitability of the approach in typical benchmarks and point out its suitability in two application scenarios: the visualization of big data sets, and the visualization of complete classifiers such as SVM.

Colloquium coordinators are Prof.dr. M. Aiello (e-mail : M.Aiello@rug.nl ) and

Prof.dr. M. Biehl (e-mail: M.Biehl@rug.nl )

http://www.rug.nl/research/jbi/news/colloquia/computerscience