Deep Reinforcement Learning and the Deadly Triad.
Hado van HasseltYotam DoronFlorian StrubMatteo HesselNicolas SonneratJoseph ModayilPublished in: CoRR (2018)
Keyphrases
- reinforcement learning
- function approximation
- state space
- reinforcement learning algorithms
- temporal difference
- dynamic programming
- optimal control
- direct policy search
- robotic control
- learning process
- control problems
- information systems
- model free
- policy search
- lung cancer
- robot control
- machine learning
- information retrieval
- perceptual aliasing
- relational reinforcement learning
- evolutionary learning
- learning agent
- learning algorithm
- reward function
- search engine
- data sets
- markov decision processes
- optimal policy
- image sequences