Colloquium Computer Science, David Nebel (Univ. of Applied Sciences, Mittweida)

Title: Learning interpretable models from Dissimilarity Data

                           

Abstract:

LVQ (Learning Vector Quantization) methods are frequently used
algorithms in the context of classication in the area of machine learning.

They combine many useful features such as sparse and interpretable models.
However, these models are only suitable for vectorial data. The existing extensions on
dissimilarity data are not longer interpretable or sparse. One method to keep
the properties is the so-called median algorithm.

In this talk I will give a brief overview of LVQ methods and their extensions
to dissimilarity data. I will name advantages and disadvantages of these methods
and motivate the introduction of so-called median methods. A brief overview
of the functionality of median methods and the advantages and disadvantages
of these will be given. It will be also shown how the median algorithms can be
applied to a wide variety of problem settings using illustrative examples.

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