Biomimicry based Learning Outcomes of Simple Cognitive Agents

We discuss a model of a complex system that could be used as an experimental virtual reality platform to study learning outcomes of a population of simple cognitive agents unable to express concepts analytically and unable to use crisp values. In modeling the cognitive agents, called “creatures”, and their learning process we pursue the route of biomimicry and steer away from formal methods and established learning algorithms, too complex for minimal creatures. The creatures use “social observational learning”, that is each creature learns from the behaviour of other creatures. The creatures may experience fear and/or desire, and are capable of evaluating if a strategy has been applied successfully and of applying this strategy again with small changes to a similar but new situation. The creatures are born as “tabula rasa”; i.e. without built-in knowledge base of their environment and as they learn they build this knowledge base. We study learning outcomes of a population of such creatures when they are learning how to safely cross various types of highways. The highways are implemented as a modified Nagel-Schreckenberg model and each creature is provided with a mechanism to reason to cross safely the highway. We present selected simulation results and their analysis.