CIDM'2011 special session on Perspectives of Dimensionality Reduction and Visual Analytics
28 Septembre 2010
Catégorie : Conférence internationale
We are please to announce a special session at IEEE CIDM'2011 on
'Perspectives of Dimensionality Reduction and Visual Analytics'
It will take place in Paris on April 11 - 15, 2011 during the Symposium on Computational Intelligence and Data Mining : http://www.ieee-ssci.org/2011/cidm-2011
Organizers: Michael Aupetit (CEA List, France), Barbara Hammer (Bielefeld University, Germany), Fabrice Rossi (Telecom ParisTech, France)
Dimensionality reduction and data visualization constitute key techniques which allow humans to intuitively scan through large volumes of complex or high-dimensional data, thereby relying on their astonishing cognitive capabilities for visual perception.
A variety of high quality nonlinear dimensionality reduction tools has recently been established including, for example, local linear embedding, Isomap, t-distributed stochastic neighbor embedding, etc. With data sets and tasks becoming more and more complex, a number of new challenges arises. The goal of the special session is to shed some light on the following issues:
- How can methods be adapted to very large data sets?
- How can the methods be evaluated properly, visualization being an inherently ill-posed task?
- How can methods be extended to manifold projection and approximate back-projection such that a deeper inspection and analysis becomes possible?
- How can the methods be combined with data mining tools (such as clustering or prediction) to get the most information from the data?
- How can the methods be adapted to widely parameterless tools suited for users outside the field?
- Paper submission due : october 31, 2010
- Notification to authors : december 15, 2010
- Camera-Ready Paper Due : January 15, 2011
Submission and information update :
Thank youMichael Aupetit