Solving Complex Machine Learning Problems with Ensemble Methods


ECML/PKDD 2013, September 27, Prague, Czech Republic

Call for papers

Scope

By combining the decisions of several different predictors, ensemble methods provide appealing solutions to challenging problems in machine learning. These include for example dealing with learning under non-standard circumstances, i.e., when large volumes of data are available for induction, or when a data stream has to be classified under the phenomenon of concept drift. Similarly, ensemble methods can be used to tackle difficult problems related to multi-label classification, feature selection, or active learning. Although research in the field of ensemble learning has grown considerably in the recent years, the specific application of ensemble methods to the problems described is still in a very early stage. There are still many open issues and there remain challenges which may require interdisciplinary approaches. This workshop aims to gather together researchers in the area of ensemble methods to present their latest work and their efforts to address difficult machine learning problems, to discuss the challenges in the field and to identify where to target our efforts as a research community. Additionally, one of the goals of the workshop is to initiate collaborations between experts in ensemble methods and non-experts. In order to achieve this objective, the workshop includes a scientific networking component, where challenging machine learning problems can be submitted for discussion.

Topics of Interest

Researchers are encouraged to submit papers focusing on how to use ensemble methods to tackle difficult machine learning problems including, but not restricted to the following topics: