Using Logical Specifications of Objectives in Multi-Objective Reinforcement Learning.
Kolby NottinghamAnand BalakrishnanJyotirmoy V. DeshmukhConnor ChristophersonDavid WingatePublished in: CoRR (2019)
Keyphrases
- multi objective
- reinforcement learning
- multiple objectives
- conflicting objectives
- multiobjective evolutionary algorithms
- evolutionary algorithm
- optimization algorithm
- pareto optimal solutions
- multi objective optimization
- multi objective evolutionary algorithms
- multiobjective optimization
- genetic algorithm
- multiobjective evolutionary algorithm
- nsga ii
- particle swarm optimization
- function approximation
- high level
- evolutionary optimization
- objective function
- single objective optimization
- multi objective optimization problems
- bi objective
- state space
- pareto optimal
- optimal policy
- multiagent evolutionary algorithm
- model free
- multi objective evolutionary
- machine learning
- learning algorithm
- formal specification
- trade off
- markov decision processes
- optimal control
- reinforcement learning algorithms
- neural network
- multi criteria
- real robot
- functional requirements
- function approximators
- genetic programming
- optimization problems
- logical operations
- simulated annealing
- robotic control
- multi agent systems
- specification language
- written in natural language