Encrypted Value Iteration and Temporal Difference Learning over Leveled Homomorphic Encryption.
Jihoon SuhTakashi TanakaPublished in: CoRR (2021)
Keyphrases
- temporal difference learning
- homomorphic encryption
- encryption scheme
- markov decision process
- policy iteration
- markov decision processes
- encrypted data
- privacy preserving
- reinforcement learning algorithms
- state space
- reinforcement learning
- data sharing
- fixed point
- standard model
- optimal policy
- temporal difference
- function approximation
- third party
- infinite horizon
- security analysis
- data hiding
- model free
- encryption algorithm
- public key
- sensitive data
- sql queries
- ciphertext
- game playing
- secret key
- lossless compression
- evaluation function
- least squares
- data storage
- dynamic programming
- read write
- partially observable markov decision processes
- initial state
- data integration
- multiresolution
- reward function
- function approximators
- optimal control
- data privacy
- average cost
- long run
- belief state