The society of mind
Intelligence without representation
Artificial Intelligence
Genetic programming: on the programming of computers by means of natural selection
Genetic programming: on the programming of computers by means of natural selection
The evolution of mental models
Advances in genetic programming
Evolutionary robotics and the radical envelope-of-noise hypothesis
Adaptive Behavior
Artificial Life: Proceedings of an Interdisciplinary Workshop on the Synthesis and Simulation of Living Systems
Concurrent Genetic Programming, Tartarus and Dancing Agents
Proceedings of the Second European Workshop on Genetic Programming
Automatic generation of object-oriented programs using genetic programming
GECCO '96 Proceedings of the 1st annual conference on Genetic and evolutionary computation
On sensor evolution in robotics
GECCO '96 Proceedings of the 1st annual conference on Genetic and evolutionary computation
Concurrent Genetic Programming, Tartarus and Dancing Agents
Proceedings of the Second European Workshop on Genetic Programming
Exploring the evolution of internal control structure using digital enzymes
Proceedings of the 14th annual conference companion on Genetic and evolutionary computation
Semantic bias in program coevolution
EuroGP'13 Proceedings of the 16th European conference on Genetic Programming
A true finite-state baseline for tartarus
Proceedings of the 15th annual conference on Genetic and evolutionary computation
An effective parse tree representation for tartarus
Proceedings of the 15th annual conference on Genetic and evolutionary computation
Hi-index | 0.00 |
Evolutionary approaches such as genetic programming have often been applied to the automatic design of controllers for autonomous agents in virtual worlds. This paper applies a multi-tree genetic programming representation to the Tartarus world. Agent-controllers are evolved whose behaviour is the emergent effect of the interleaved evaluation of the program trees. Agents with good fitness and of very low complexity are evolved, and it is found that this technique evolves agents that exploit the characteristics of the runtime scheduler to provide an implicit rather than explicit form of state in the form of a "fixed dance".