Adaptive Behavior current issue

A Computational Model of Social-Learning Mechanisms

Lopes, M., Melo, F. S., Kenward, B., Santos-Victor, J. — Thu, 19 Nov 2009 04:47:11 PST

In this article we propose a computational model that describes how observed behavior can influence an observer’s own behavior, including the acquisition of new task descriptions. The sources of influence on our model’s behavior are: beliefs about the world’s possible states and actions causing transitions between them; baseline preferences for certain actions; a variable tendency to infer and share goals in observed behavior; and a variable tendency to act efficiently to reach rewarding states. Acting on these premises, our model is able to replicate key empirical studies of social learning in children and chimpanzees. We demonstrate how a simple artificial system can account for a variety of biological social transfer phenomena, such as goal-inference and over-imitation, by taking into account action constraints and incomplete knowledge about the world dynamics.

Incremental Learning and Memory Consolidation of Whole Body Human Motion Primitives

Kulic, D., Nakamura, Y. — Thu, 19 Nov 2009 04:47:11 PST

The ability to learn during continuous and on-line observation would be advantageous for humanoid robots, as it would enable them to learn during co-location and interaction in the human environment. However, when motions are being learned and clustered on-line, there is a trade-off between classification accuracy and the number of training examples, resulting in potential misclassifications both at the motion and hierarchy formation level. This article presents an approach enabling fast on-line incremental learning, combined with an incremental memory consolidation process correcting initial misclassifications and errors in organization, to improve the stability and accuracy of the learned motions, analogous to the memory consolidation process following motor learning observed in humans. Following initial organization, motions are randomly selected for reclassification, at both low and high levels of the hierarchy. If a better reclassification is found, the knowledge structure is reorganized to comply. The approach is validated during incremental acquisition of a motion database containing a variety of full body motions.¹

Pattern-Oriented Modeling of Commons Dilemma Experiments

Janssen, M. A., Radtke, N. P., Lee, A. — Thu, 19 Nov 2009 04:47:11 PST

A major challenge in the development of computational models of collective behavior is the empirical validation. Experimental data from a spatially explicit dynamic commons dilemma experiment is used to empirically ground an agent-based model. Three distinct patterns are identified in the data. Two naïve models, random walk and greedy agents, do not produce data that match the patterns. A more comprehensive model is presented that explains how participants make movement and harvest decisions. Using pattern-oriented modeling the parameter space is explored to identify the parameter combinations that meet the three identified patterns. Less than 0.1% of the parameter combinations meet all the patterns. These parameter settings were used to successfully predict the patterns of a new set of experiments.

Partner Search Heuristics in the Lab: Stability of Matchings Under Various Preference Structures

Eriksson, K., Strimling, P. — Thu, 19 Nov 2009 04:47:11 PST

When agents search for partners, the outcome is a matching. K. Eriksson and O. Häggström (2008) defined a measure of instability of matchings and proved that under a certain partner search heuristic, outcomes are likely to have low instability. They also showed that with regards to stability, the preference structure known as common preferences lie somewhere in between the extreme cases of homotypic and antithetical preferences. Following up on this theoretical work, we let human subjects search for a good partner in a computer game where preferences were set to be either common, homotypic, or antithetical. We find that total search effort and instability of the outcome vary in the predicted ways with the preference structure and the number of agents. A set of simulations show that these results are consistent with a model where agents use a simple search heuristic with a slight possibility of error.

Levels and Types of Action Selection: The Action Selection Soup

Ozturk, P. — Thu, 19 Nov 2009 04:47:11 PST

Action selection (AS) is defined as the process where an action is selected among a number of alternatives. This definition, however, does not sufficiently describe what an action is. What is the unit of selection in the first place? We maintain that the artificial intelligence (AI) accounts of AS typically mix and merge two AS situations that indeed are qualitatively different. Most of the accounts actually deal only with one type of AS but purport to cover both types of AS. We propose three dimensions along which the commonalities and the differences between various AS accounts can be analyzed, and use these for a preliminary conceptualization of what we call a two-system action selection account. In particular, we identify two qualitatively different AS situations whose architectures, we suggest, can be designed inspired by neuroscience models of the basal ganglia (BG) and the cerebellum, respectively.

Thanks to Reviewers

Thu, 19 Nov 2009 04:47:11 PST