RewardsOfSum: Exploring Reinforcement Learning Rewards for Summarisation.
Jacob ParnellInigo Jauregi UnanueMassimo PiccardiPublished in: SPNLP@ACL-IJCNLP (2021)
Keyphrases
- reinforcement learning
- markov decision processes
- function approximation
- reinforcement learning algorithms
- model free
- state space
- transfer learning
- reward function
- temporal difference
- learning algorithm
- reward shaping
- hidden state
- reinforcement learning methods
- control policy
- function approximators
- learning classifier systems
- optimal policy
- least squares
- machine learning
- fully observable
- partial observability
- multiarmed bandit
- learning capabilities
- learning problems
- monte carlo
- dynamic programming
- information retrieval