Inferring Continuous Treatment Doses from Historical Data via Model-Based Entropy-Regularized Reinforcement Learning.
Jianxun WangDavid RobertsAndinet EnquobahriePublished in: ACML (2020)
Keyphrases
- historical data
- reinforcement learning
- model free
- stock price
- action space
- information theoretic
- data mining techniques
- demand forecasting
- stream data
- function approximation
- continuous state and action spaces
- information theory
- mutual information
- continuous state spaces
- temporal difference
- predictive model
- reinforcement learning algorithms
- learning algorithm
- machine learning
- markov decision processes
- data mining technology
- state space
- dynamic programming
- continuous state
- learning process
- power plant
- customer behavior
- data analysis
- fitted q iteration