Agents that Learn to Behave in Multi-Agent Simulations

J. Dávila and M. Uzcátegui (Venezuela)


ILP, Multi-agent simulation, logic-based agents


This paper illustrates the use of Inductive Logic Program ming to program agents that learn rules of behaviour from simulated histories of their embedding systems. We have shown how a ILP system can be used to learn rules in a representation very close to the one used to guide the sim ulation of a multi-agent system. This establishes the feasi bility of embedding (resource-bounded) learners as agents that take part in simulating a complex system.

Important Links:

Go Back