During the last decades, the machine learning and related communities have conducted numerous researches with the purpose of improving the performance of a single classifier through combining several classifiers generated from one or more learning algorithms. These systems are called multiple classifier systems, ensemble of classifiers, or ensemble methods.

The strategy in ensemble methods is to create a set of accurate and diverse classifiers and combine their outputs such that the combination outperforms every one of the single classifiers. Studies in the ensemble field have typically focused on generating the ensemble members applying one or various learning algorithms and combining their outputs using a mathematical function or a meta-classifier. Bagging, Boosting and

Stacking are the most popular and most representative techniques of multiple classification systems. Such methods have been successfully applied to a wide range of real problems and have been the germ of a large number of new architectures and proposals.

In addition, ensemble methods have been used in evolving frameworks with promising results. The objective of this special session is to provide a forum for the discussion of recent research (both theoretical and practical) focused on design of ensemble methods in evolving frameworks and their application in different fields, such as: finance, robotics, bioinformatics, healthcare, etc.

María Paz Sesmero Lorente, Carlos III University of Madrid, Spain

José Antonio Iglesias Martínez, Carlos III University of Madrid, Spain