Deep Reinforcement Learning Policies Learn Shared Adversarial Features Across MDPs.
Ezgi KorkmazPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- optimal policy
- markov decision processes
- policy search
- state space
- markov decision process
- multi agent
- reinforcement learning agents
- feature set
- feature extraction
- markov decision problems
- feature vectors
- dynamic programming
- reward function
- control policies
- decision problems
- learning agent
- function approximation
- markov games