An Efficient Hardware Implementation of Reinforcement Learning: The Q-Learning Algorithm.
Sergio SpanòGian Carlo CardarilliLuca Di NunzioRocco FazzolariDaniele GiardinoMarco MattaAlberto NannarelliMarco RePublished in: IEEE Access (2019)
Keyphrases
- hardware implementation
- reinforcement learning
- learning algorithm
- reinforcement learning algorithms
- signal processing
- function approximation
- efficient implementation
- hardware architecture
- software implementation
- learning problems
- machine learning
- machine learning algorithms
- dedicated hardware
- fpga implementation
- supervised learning
- hardware design
- memory management
- state space
- pipeline architecture
- field programmable gate array
- markov decision processes
- active learning
- learning process
- multi agent
- parallel architecture
- fitted q iteration
- model free
- image processing algorithms
- temporal difference
- real time
- learning agent
- high speed
- pattern recognition
- computer vision
- data mining