Tunable Behaviours in Sequential Social Dilemmas using Multi-Objective Reinforcement Learning.

David O'Callaghan Patrick Mannion

Published in: AAMAS (2021)

Keyphrases

multi objective
reinforcement learning
evolutionary algorithm
multi objective optimization
optimization algorithm
multiobjective optimization
particle swarm optimization
objective function
conflicting objectives
genetic algorithm

nsga ii
function approximation
state space
evolutionary optimization
optimal policy
multi objective optimization problems
bi objective
temporal difference
multiple objectives
learning algorithm

learning problems
multiobjective evolutionary algorithm
learning process
reinforcement learning algorithms
model free
multi criteria
trade off
optimal control
evolutionary computation
pareto optimal

markov decision processes
optimization problems
markov decision process
multi agent reinforcement learning
multi agent systems
transition model
optimum design
multi objective evolutionary
multi objective genetic algorithms
multiagent evolutionary algorithm