An Adaptive Deep RL Method for Non-Stationary Environments with Piecewise Stable Context.
Xiaoyu ChenXiangming ZhuYufeng ZhengPushi ZhangLi ZhaoWenxue ChengPeng ChengYongqiang XiongTao QinJianyu ChenTie-Yan LiuPublished in: CoRR (2022)
Keyphrases
- computational cost
- significant improvement
- high precision
- reinforcement learning
- objective function
- high accuracy
- cost function
- experimental evaluation
- synthetic data
- detection method
- learning algorithm
- dynamic programming
- computational complexity
- prior knowledge
- evolutionary algorithm
- edge detection
- input data
- model selection
- non stationary
- pairwise
- preprocessing
- synthetic and real images
- model free