Call for Papers and Datasets
Data and Machine Learning Approaches with Multiple Views
In conjunction with ECML-PKDD’19
Recent years have witnessed new frameworks/algorithms able to deal with multiple views, such as Multiple Kernel Learning, Boosting, Co-regularized, Deep approaches. Such algorithms come from the Machine Learning community and find applications in many different areas, such as Multimedia Indexing, Computer Vision, Bio-informatics, Neuro-imaging… Multiview learning, naturally enough, emphasises the potential benefits of learning through collaboration with multiple sources of data (e.g. video document can be described through images, sound, motion, text).
Depending on the context, this issue of learning from multiple descriptions of data goes under the name of multiview learning (machine learning, computer vision), multimodality fusion (multimedia), among others.
This workshop is the opportunity to bring together theoretical and applicative communities around multiview learning, which could lead to significant contributions and exchanges between Machine Learning and natural fields of applications such as biology, computer vision, marketing, ecology, health, computer vision, etc.
The workshop aims at bringing together people interested with multi-view learning, both from dataset providers to researchers in machine learning. Such a way, researchers could easily have the opportunity to inspect the reality of some true learning problems related to multi-view
learning, meanwhile providers of natural multi-viewed data could get aware of the many current or potential solutions to address their learning tasks.
This workshop will dedicate the morning to talks about multi-view theory and algorithms and talks about multi-view real datasets/tasks. In the afternoon, a hackhaton on one of the selected dataset will be organized where attendants are expected to produce a team work including datasets providers for drawing relevant solutions. The call for contributions is then twofolds:
- Call for papers (extended abstracts) on multi-view learning methods, theory, and applications
- Call for true multiview datasets (for presentation of the dataset and maybe for the hackhaton)
Please note that at least one author of each accepted paper should register for the ECML/PKDD conference.
Topics of interests
The main objectives of this workshop are to 1) introduce recent development in machine learning with multiview setting, 2) focus on various problematics in any field where such a setting arises and
3) offer new directions and discuss about open questions that appear.
In particular, the following topics specific to multi-view learning are relevant:
• Diversity / Complementarity / (Dis-)agreement between views
• Missing data or views / Noisy data or views / Noisy annotations
• Multiview for Large-scale / Big data
• Multiview for small data
• Relevant losses and theory for multiview learning
• Scaling multiview approaches
• Multiview and Ranking / Learn with imbalanced data set
• Variables and views selection
• Representation Learning (deep or not) with multiple views
• Multiview for domain adaptation / transfer learning / optimal transport