A Deep Reinforcement Learning Method for Freight Train Driving Based on Domain Knowledge and Mass Estimation Network.
Wenlong FengWei DongShouchao ZhaiGuohua ZhangXinya SunYindong JiPublished in: ICMLT (2021)
Keyphrases
- domain knowledge
- reinforcement learning
- detection method
- high accuracy
- preprocessing
- significant improvement
- cost function
- pairwise
- network model
- transportation networks
- dynamic programming
- edge detection
- high precision
- clustering method
- estimation process
- estimation accuracy
- density estimation
- segmentation method
- machine learning
- objective function
- prior knowledge