Transfer Learning (TL) aims to transfer knowledge acquired in one problem, the source problem, onto another problem, the target problem, dispensing with the bottom-up construction of the target model. The TL approach has gained significant interest in the Machine Learning (ML) community since it paves the way to devise intelligent learning models that can easily be tailored to many different domains of applicability.
The following aspects have recently contributed to the emergence of TL:
• Generalization Theory: TL often produces algorithms with good generalization capability for different problems;
• Efficient TL algorithms: TL provides learning models that can be applied with far less computational effort than standard ML methods;
• Unlabeled data: TL can be advantageous since unlabeled data can have severe implications in some fields of research, such as in the biomedical field.

The topics of this special session are:
• Big Data with Deep Neural Networks;
• Generalization Bounds;
• Domain Adaptation or Covariate Shift;
• Algorithms for TL;
• New advancements in TL;
• Real-world applications.

Web: http://iwann.ugr.es/2015
http://iwann.ugr.es/2015/iwann15.php?menu=contributions&sub=specialsesions

Submission process: The submission process follows the same guidelines as any other paper on IWANN. Please refer to Authors Information in the IWANN webpage. When submitting choose Transfer Learning as the Category.

Organizers:
Luís M. Silva - Dep. of Mathematics, University of Aveiro, Portugal - lmas@ua.pt
Jorge M. Santos - Dep. of Mathematics, Engineering School of the Polytechnic of Porto, Portugal -jms@isep.ipp.pt

For further assistance please contact one of the organizers.