Local-utopia policy selection for multi-objective reinforcement learning.
Simone ParisiAlexander BlankTobias ViernickelJan PetersPublished in: SSCI (2016)
Keyphrases
- multi objective
- reinforcement learning
- optimal policy
- policy search
- evolutionary algorithm
- optimization algorithm
- multi objective optimization
- genetic algorithm
- particle swarm optimization
- multiobjective optimization
- action selection
- control policy
- markov decision process
- multiple objectives
- partially observable environments
- policy gradient
- function approximation
- temporal difference
- multi agent
- function approximators
- conflicting objectives
- multi objective optimization problems
- approximate dynamic programming
- policy evaluation
- state dependent
- learning algorithm
- control policies
- reinforcement learning problems
- state and action spaces
- reward function
- markov decision problems
- neural network
- pareto optimal
- selection algorithm
- markov decision processes
- objective function
- machine learning
- state action
- reinforcement learning algorithms
- actor critic
- inverse reinforcement learning