FA3C: FPGA-Accelerated Deep Reinforcement Learning.
Hyungmin ChoPyeongseok OhJiyoung ParkWookeun JungJaejin LeePublished in: ASPLOS (2019)
Keyphrases
- reinforcement learning
- hardware implementation
- state space
- reinforcement learning algorithms
- learning algorithm
- function approximation
- real time image processing
- optimal policy
- high speed
- field programmable gate array
- signal processing
- hardware design
- deep learning
- low cost
- real time
- machine learning
- fpga implementation
- multi agent reinforcement learning
- markov decision processes
- temporal difference
- model free
- transfer learning
- function approximators
- single chip
- learning process
- robotic control
- digital signal
- dynamic programming
- verilog hdl
- fpga hardware
- systolic array
- policy search
- multi agent
- computer systems
- learning capabilities
- reward function