Connectionists: special session on icremental learning at ESANN'2016

[Barbara Hammer]

*CALL FOR PAPERS*

We encourage submissions to the special session

*Incremental learning algorithms and applications*

organized by Alexander Gepperth, ENSTA ParisTech (France), Barbara 
Hammer, Univ. Bielefeld (Germany)

at ESANN 16 - European Symposium on Artificial Neural Networks and 
Machine Learning
https://www.elen.ucl.ac.be/esann/
which will take place from 27-29 April in beautiful Bruges, Belgium.

This special sessions invites contributions from researchers in machine 
learning who develop incremental learning algorithms, as well as from 
researchers in applied fields, particularly robotics, who apply such 
algorithms to obtain superior performance or usability. The definition 
of incremental learning algorithms is notoriously unclear: It may 
include, but is not restricted to, algorithms that:

• avoid catastrophic forgetting when retrained
• can deal with non-stationary statistics: concept drift or concept shift
• can treat "streaming" data and see each example only one
• try to emulate biological architectures of life-long learning

A particular focus lies on algorithms that rest efficient (w.r.t. time 
and memory) when faced with huge amounts of data, or data with large 
dimensionality, or both.

The deadline for submissions is November 20, 2015 via the conference portal
https://www.elen.ucl.ac.be/esann/

-- 
Prof. Dr. Barbara Hammer
CITEC centre of excellence
Bielefeld University
D-33594 Bielefeld
Phone: +49 521 / 106 12115
Fax:   +49 521 / 106 12181