Skip to main navigationSkip to main content
The University of Southampton
CORMSIS Centre for Operational Research, Management Sciences and Information Systems

Classifier Monitoring and Updating from Unlabelled Data Event

16:00 - 17:00
25 February 2016
Room 3041, Building 2, Southampton Business School

For more information regarding this event, please email Dr Yuan Huang at .

Event details

Abstract: In a dynamic world, the distributions on which prediction models are based are not stationary but change over time. The question whether, when and how to incorporate effects of such concept drift into a prediction model is highly relevant. A classic approach is to monitor the predictive performance on labelled data. However, such data is often scarce and delayed, and the predictive performance is not necessarily a good indicator of concept drift. Recently, techniques have been proposed that either monitor the data distributions directly, extrapolate continuous change patterns, or even incorporate information from yet unlabelled data. In this talk, I will review those techniques and propose an approach that combines monitoring and the prediction of gradual, continuous changes in the data by applying drift mining techniques to unlabelled data. It aims to assess the type of drift in the data and monitors in particular for unexpected changes. Such changes denote changes that are not predictable by extrapolating change patterns. The approach monitors the distribution of recent, unlabelled data and delayed, labelled data to indicate moments and regions of unexpected change in the spatio-temporal distribution.

Speaker information

Dr Georg Krempl,Otto-von-Guericke-University Magdeburg, Germany ,Bio: Dr Georg Krempl is a postdoc researcher in the Knowledge Management and Discovery Lab at the Otto-von-Guericke-University Magdeburg, Germany. He received his doctorate in 2011 from University of Graz, Austria, where he worked on multiple classifier systems and adaptive classification algorithms for credit scoring. His current research is focused on adaptive learning algorithms for evolving data streams, where he addresses feedback-associated challenges such as delayed labels/verification latency, temporal transfer learning or active, cost-minimizing selection of information. He has published several journal and proceedings papers and co-organized tutorials and a workshop at international data-mining conferences (ECML PKDD 2012, PAKDD 2013, ECML PKDD 2013, iKnow 2015).

Privacy Settings