Objectives
Anticipatory behavior is a mechanism, or a behavior, that does not only depend on the past and present but also on predictions, expectations, or beliefs about the future.
The aim of this workshop is to bring together researchers that are interested in such anticipatory mechanisms and essentially anticipatory adaptive behavior. It is aimed for an interdisciplinary gathering that brings together researchers from distinct areas so as to discuss the different guises that anticipation takes in these different perspectives. After a very successful workshop on Anticipatory Behavior at SAB 2002, resulting in the Springer-Verlag LNCS 2684 State-of-the-Art survey named after the workshop, we are interested in building upon that initiative emphasizing the exploitation of anticipations in learning and planning.
Previous work on anticipatory behavior has concentrated more on the learning of environmental models, especially online generalized model learning. Up to now though, exploitation of the model has been done mainly to show that exploitation is possible or that an appropriate model exists in the first place. Only very few applications exist that show the utility of the model for the simulation of anticipatory processes and consequent adaptive behavior. However, the exploitation of the model and the interaction of learning and behavior by the means of the model is the most promising and important area for future research.
Essential Questions
- How can anticipations be exploited to improve planning?
- How can anticipations be exploited to direct or speed-up learning?
- Which social agent interactions can be improved by the means of anticipatory behavior?
- How do anticipations influence attention?
- How can anticipations improve the adaptive behavior of an artificial learning system?
- How can anticipatory adaptive behavior be implemented in an artificial learning system?
- How does an incomplete predictive model influence anticipatory behavior?
- How do anticipations guide further model learning?
- Can anticipations be used for the detection of special environmental properties?
- What role do anticipations play for the implementation of motivations and emotions?
- What is the trade-off between simple bottom-up stimulus-response driven behavior and more top-down anticipatory driven behavior?
- In what respect does anticipation mediate between low-level environmental processing and more complex cognitive simulation?
- Which anticipatory mechanisms can be distinguished? Which are the benefits/drawbacks of the different mechanisms?