Safety Optimized Reinforcement Learning via Multi-Objective Policy Optimization.
Homayoun HonariMehran Ghafarian TamiziHomayoun NajjaranPublished in: CoRR (2024)
Keyphrases
- rl algorithms
- reinforcement learning
- multi objective
- optimization algorithm
- optimal control
- multiple objectives
- evolutionary algorithm
- conflicting objectives
- optimal policy
- evolutionary optimization
- optimum design
- multiobjective optimization
- function approximation
- average reward
- multi objective optimization
- engineering design problems
- optimization problems
- optimization method
- reinforcement learning algorithms
- policy search
- function approximators
- objective function
- genetic algorithm
- particle swarm optimization
- particle swarm
- action selection
- pareto optimal
- global optimization
- dynamic programming
- machine learning
- temporal difference
- differential evolution
- multi objective evolutionary algorithms
- multi agent
- bi objective
- infinite horizon
- partially observable environments
- multi objective optimization problems
- multi objective evolutionary
- continuous state
- control policies
- long run
- control policy
- estimation of distribution algorithms
- policy iteration
- partially observable markov decision processes
- partially observable