Multi-objective optimization perspectives on reinforcement learning algorithms using reward vectors.
Madalina M. DruganPublished in: ESANN (2015)
Keyphrases
- multi objective optimization
- reinforcement learning algorithms
- reinforcement learning
- eligibility traces
- reward function
- multi objective
- total reward
- partially observable environments
- evolutionary algorithm
- markov decision processes
- model free
- state space
- reinforcement learning problems
- policy gradient
- learning agent
- function approximation
- reinforcement learning methods
- genetic algorithm
- pareto optimal
- multi objective optimization problems
- temporal difference
- learning algorithm
- optimal policy
- nsga ii
- reward shaping
- differential evolution
- dynamic programming
- bi objective
- multiple objectives
- average reward
- machine learning
- multi objective genetic algorithm
- multiple agents
- inverse reinforcement learning
- markov decision process
- multiobjective optimization
- artificial neural networks
- multi objective genetic algorithms
- fitness function
- simulated annealing
- evolutionary computation
- multi agent
- decision making
- optimal control