Black-Box Reward Attacks Against Deep Reinforcement Learning Based on Successor Representation.
Kanting CaiXiangbin ZhuZhao-Long HuPublished in: IEEE Access (2022)
Keyphrases
- black box
- reinforcement learning
- black boxes
- white box
- function approximation
- markov decision processes
- integration testing
- databases
- test cases
- rule extraction
- white box testing
- database
- hybrid systems
- state transition
- metadata
- reward function
- reinforcement learning algorithms
- temporal difference
- optimal policy
- computational intelligence
- eligibility traces
- dynamic programming