Special Session Call for
Dynamic learning in non-stationary environments
AIMS AND SCOPE
The computerization of many life activities and the advances in data collection and storage technology lead to obtaining mountains of data. These data capture information about various phenomena and process behaviour and are rarely of direct benefit. Thus, a set of techniques and tools should be used to extract useful information for decision support, prediction, exploration and understanding of phenomena governing the data sources.
Learning methods use historic data points about a process past behavior to build a predictor (classifier, regression model), which is used to predict the process future behaviour. However, the predictor needs to adjust itself (called self-correction or adaptation) as new events happen or new conditions/system states occur (e.g., during on-line operations). The goal is to ensure an accurate prediction of the process behaviour according to the changes in the new incoming data characteristics. This requires a continuous learning over a long period of time, with the ability to evolve new structural components on demand and to forget data that became obsolete and useless. Incremental and sequential learning are essential concepts in order to avoid time-intensive re-training phases as well as account for the systems dynamics/changing data characteristics, with low computational effort and virtual memory usage (enhancing the on-line performance). This is because data is processed in sample-wise and single-pass manner.
It is important that the updates of model parameters and dynamic changes in structural components are achieved without a “catastrophic forgetting”. Therefore, a balance between continuous learning and “forgetting” is necessary in order to deal with non-stationary environments.
This special session looks to bring together and discuss efficient machine learning techniques, methods and tools able to manage, exploit and interpret correctly the increasing amount of data in environments that are continuously changing. The goal is to build effective models for predicting the future system behavior, which are able to tackle and govern the high variability of complex non-stationary systems.
This session solicits original research papers including but not limited to the following topics:
· Incremental learning methods
· Adaptive, life-long and sequential learning
· On-line classification and regression methods
· Evolving structural components and system modelling
· Incremental/evolving un-supervised methods
· Incremental/on-line dimensionality reduction methods
· Concepts to address drifts and shifts in data streams (weighting, gradual forgetting, etc.)
· On-line/incremental active and semi-supervised learning
· On-line human-machine interaction and the incorporation of background knowledge
· Adaptive data pre-processing and knowledge discovery
· Applications of dynamic/on-line/incremental learning for:
o On-line quality control systems
o Fault detection and isolation
o Huge data bases
o Web applications
o Decision Support Systems
o And many more ….
Paper Submission Deadline: July 15, 2010
Notification of acceptance: September 7, 2010
Camera-ready papers & Pre-registration: October 1, 2010
The ICMLA Conference: December 12-14, 2010
This special session will be held as part of the ICMLA’10 conference. Authors should submit papers through the main conference submission website. Papers must correspond to the requirements detailed in the instructions to authors. All conference submissions will be handled electronically. Detailed instructions for submitting the papers are provided on the conference home page at:
Accepted papers should be presented by one of the authors in order to be published in the conference proceeding. If you have any questions, do not hesitate to direct your questions to the session organizers.
· Moamar Sayed-Mouchaweh, University of Reims, France, firstname.lastname@example.org
· Edwin Lughofer, University of Linz, Austria, email@example.com