AISB 2011 Symposium on: Computational Models of Cognitive Development

Programme (April 5th)

 9:00 - 10:00 Plenary speaker: Alan Baddeley
    						
10:00 - 10:30 Coffee break 		

----------------------------------------

10:30 - 12:30 Symposium session 1
                Setting the Agenda: Brief perspectives on how to
                build a complete model of a developing system

10:30 - 10:54 Mark Lee (Aberystwyth)
                Perspective: Psychology and Developing Robots

10:54 - 11:18 Honghai Liu (Portsmouth)
                Perspective:
                Hand Skill Transfer: a Cognitive Challenge

11:18 - 11:42 Norbert Krueger (Southern Denmark)
                Perspective:
                Vision for Cognition: Prior knowledge and Development

11:42 - 12:06 Mark Steedman (Edinburgh)
                Perspective: Where does compositionality come from?

12:06 - 12:30 John Alexander (Aberdeen)
                Perspective: Hierarchical Knowledge Structures

----------------------------------------
    						
12:30 - 14:00 Lunch 		
    					
14:00 - 15:00 Plenary speaker: Mark Steedman
    						
----------------------------------------

15:00 - 16:30 Symposium session 2 		
                Psychology focus
                (25 minute talks, 5 minute discussions;
                 Chair: Frank Guerin)

15:00 - 15:30 Katherine E. Twomey and Jessica S. Horst
                All Things Considered: Dynamic Field Theory Captures
                Effect of Categories on Children's Word Learning

15:30 - 16:00 Anne S. Warlaumont, Gert Westermann,
                and D. Kimbrough Oller
                Self-production Facilitates and Adult Input Interferes
                in a Network Model of Infant Vowel Imitation

16:00 - 16:30 Alexandra Varga and Michiel van Lambalgen
                Infants' Flexible Closed World Reasoning and
                Imitation-for-Learning
    						
----------------------------------------

16:30 - 17:00 Coffee break 		
    						
----------------------------------------

17:00 - 18:30 Symposium session 3 				
                Artificial Intelligence focus
                (25 minute talks, 5 minute discussions;
                 Chair: John Alexander)

17:00 - 17:30 Norbert Krueger, Mila Popovic, Leon Bodenhagen,
                Dirk Kraft and Frank Guerin
                Grasp Learning by Means of Developing Sensorimotor
                Schemas and Generic World Knowledge

17:30 - 18:00 Michael Miller
                Piagetian Autonomous Modeller

18:00 - 18:30 Timothy A. Furze and Brandon Bennett
                Using the Principals of Classical Conditioning to Learn
                Event Sequences
    						
----------------------------------------

20:00  	      Convention dinner

Call for papers

Since the early days of Artificial Intelligence there has been interest in the idea of building a "baby intelligence" which could learn for itself (Turing, 1950); yet our current understanding of how artificial or biological systems can develop cognitively is still relatively poor, especially when considering issues such as ongoing emergence in autonomous cognitive development (Prince et al., 2005).

There is by now a reasonable body of work in computational models of cognitive development, yet the field as a whole is currently in a fragmentary state. There exist models of particular isolated episodes in development, for example learning to reach (Schlesinger, Parisi, and Langer, 2000), or the A-not-B error (e.g. Thelen, Schöner, Scheier, and Smith, 2001), or the balance scale problem (Shultz, Mareschal, and Schmidt, 1994); however there is not yet a single integrated longitudinal model which could put multiple episodes together and account for a sequence of developmental episodes. There also exist models of grand theories of development, such as models of Piaget's theory (Drescher, 1991; Stojanov, 2001; Chaput, 2004), yet it is not clear how generally applicable these models are, beyond the domains on which they have been tested. Furthermore, it is increasingly being recognised that models of cognitive development cannot exist in isolation from perceptual and motor development.

In parallel with the above works there are numerous works in the AI and robotics communities which tackle issues central to development, such as intrinsic motivation (the topic of a recent special issue of the IEEE TAMD journal), or learning of action hierarchies (e.g. Mugan and Kuipers, 2009; Hart, Sen, Grupen, 2008); the AI techniques used in these works go beyond what has typically been used in models of cognitive development in biological systems, and may well prove to be valuable for modelling aspects of development which current models cannot account for. In the other direction, those engaged in AI and developmental robotics may well benefit from greater connections with scientists researching cognitive development in biological systems.

This symposium aims to bring together researchers engaged in these related areas of work in order to facilitate new collaborations which could lead towards research connecting up the existing fragmentary models and techniques. The long-term aim of this collaboration is to advance our understanding of how cognitive development can occur, in biological or artificial systems; a complete understanding would need to provide an integrated account for developments at different timescales, and in different domains, and especially for how successive developments build on each other.

Accepted papers are published in a volume by the SSAISB, and receive an ISBN number. AISB does not demand exclusive copyright, and so does not preclude later publication of the same material.

We welcome submissions from anyone interested in themes including but not limited to those below, and especially encourage interdisciplinary submissions which link previously unassociated fields.

- Connectionist models of cognitive development
- Dynamic Systems models of cognitive development
- Models of stage transition
- Models of developmental theories (e.g. Piaget's or Fischer's)
- Models integrating psychological and neuroscientific evidence
- Biologically plausible developmental robotics
- The role of social learning and imitation in development
- Models of intrinsic motivation and value systems
- Sensorimotor learning and sensorimotor abstraction
- Schema learning mechanisms
- Bootstrap learning with limited innate knowledge
- The role of play in cognitive development
- Architectures for developmental systems
- Ongoing emergence and cumulative learning
- Autonomous development of new representations
- Hierarchical learning of skills and/or representations
- Learning Gibsonian affordances

Programme Committee

How to Submit

Location and Travel