In the last years there has been an increase of activity in the areas of domain adaptation, transfer and multi-task learning. All born as solutions to better exploit the available data at training time and often moved by the need to deal with a reduced amount of information, these three topics grew fast in several directions and have multiple applications. Today new open research questions present the challenge. On one side, the literature is missing a joint theoretical framework over all of them, replaced instead by many theoretical formulations model regimes that are rarely used in practice (e.g. adaptive methods that store all the source samples). On the other, in the “big data” era, existing methods should be extended to manage large amount of data that do not lack anymore in size but may lack in quality or may continuously change over time. 

This special topic is intended to to gather contributions that indicate new directions, innovative views and to serve as an outlet for recent advances in learning in such environments. We welcome both theoretical advances in this field as well as detailed reports on applications.

Topics of interest include:

• Learning the task similarities/dissimilarities from large amount of data
• Regularization strategies in multitask learning
• Domain adaptation and dataset bias on large data collections
• Deep Learning on domain adaptation, transfer and multi-task applications
• Incremental, online and active transfer for open-ended learning
• Innovative adaptive procedures with applications e.g. in computer vision or computational biology
• Domain adaptation theory
• Large number of tasks, small number of examples setting in multitask learning
• Reinforcement learning and adaption
• Applications of multitask learning to natural language processing tasks such as machine translation and syntactic or semantic parsing.  

• Submission: 15 April 2015  EXTENDED DEADLINE: 15 May 2015
• Decision: 15 August 2015   First review round completed
  
• Final Version and Resubmissions Due: 6 weeks after decision communication