Preference-Controlled Multi-Objective Reinforcement Learning for Conditional Text Generation.
Wenqing ChenJidong TianCaoyun FanYitian LiHao HeYaohui JinPublished in: AAAI (2023)
Keyphrases
- multi objective
- text generation
- reinforcement learning
- natural language generation
- multi objective optimization
- evolutionary algorithm
- multi criteria
- multiple criteria
- optimization algorithm
- natural language
- genetic algorithm
- nsga ii
- cp nets
- function approximation
- multiple objectives
- conflicting objectives
- particle swarm optimization
- theorem prover
- state space
- objective function
- multi objective optimization problems
- multi agent
- model free
- multi attribute
- temporal difference
- multi objective evolutionary
- reinforcement learning algorithms
- preference elicitation
- pareto optimal
- conditional probabilities
- dynamic programming
- learning process
- preference relations
- bi objective
- differential evolution
- markov decision processes
- learning algorithm