Safety Optimized Reinforcement Learning via Multi-Objective Policy Optimization.
Homayoun HonariMehran Ghafarian TamiziHomayoun NajjaranPublished in: ICRA (2024)
Keyphrases
- multi objective
- optimization algorithm
- reinforcement learning
- multiple objectives
- optimal policy
- evolutionary algorithm
- evolutionary optimization
- conflicting objectives
- optimum design
- multiobjective optimization
- multi objective optimization
- multi objective evolutionary algorithms
- particle swarm
- markov decision process
- action selection
- optimization method
- policy search
- engineering design problems
- markov decision processes
- partially observable environments
- pareto optimal
- genetic algorithm
- optimization problems
- objective function
- learning algorithm
- function approximation
- global optimization
- particle swarm optimization
- state and action spaces
- control policies
- state space
- estimation of distribution algorithms
- state action
- policy gradient
- multi objective evolutionary
- reinforcement learning algorithms
- reward function
- multi agent
- average reward
- control policy
- bi objective
- partially observable
- policy evaluation
- approximate dynamic programming
- actor critic
- reinforcement learning problems
- model free
- linear programming
- mobile robot
- neural network