Designing Non-greedy Reinforcement Learning Agents with Diminishing Reward Shaping.
Fan-Yun SunYen-Yu ChangYueh-Hua WuShou-De LinPublished in: AIES (2018)
Keyphrases
- reinforcement learning agents
- reward shaping
- reinforcement learning
- dynamic environments
- state abstraction
- function approximation
- search algorithm
- dynamic programming
- reinforcement learning algorithms
- transfer learning
- optimal control
- state space
- multi agent
- optimal policy
- multi agent environments
- feature selection
- learning algorithm
- complex domains
- markov decision problems
- cooperative