Behavior-Targeted Attack on Reinforcement Learning with Limited Access to Victim's Policy.
Shojiro YamabeKazuto FukuchiRyoma SendaJun SakumaPublished in: CoRR (2024)
Keyphrases
- reinforcement learning
- optimal policy
- policy search
- action selection
- approximate dynamic programming
- control policy
- reinforcement learning algorithms
- state and action spaces
- state space
- reinforcement learning problems
- cache management
- policy iteration
- selective perception
- policy evaluation
- markov decision process
- partially observable markov decision processes
- countermeasures
- function approximation
- action space
- state dependent
- access control
- markov decision processes
- actor critic
- model free
- machine learning
- partially observable environments
- multi agent
- markov decision problems
- control policies
- reward function
- function approximators
- neural network