Matches in SemOpenAlex for { <https://semopenalex.org/work/W3175606556> ?p ?o ?g. }
Showing items 1 to 87 of
87
with 100 items per page.
- W3175606556 abstract "Reinforcement Learning (RL) represents the machine learning method that has come closest to showing human-like learning. While Deep RL is becoming increasingly popular for complex applications such as AI-based gaming, it has a high implementation cost in terms of both power and latency. Q-Learning, on the other hand, is a much simpler method that makes it more feasible for implementation on resource-constrained embedded systems for control and navigation. However, the optimal policy search in Q-Learning is a compute-intensive and inherently sequential process and a software-only implementation may not be able to satisfy the latency and throughput constraints of such applications. To this end, we propose a novel accelerator design with multiple design trade-offs for implementing Q-Learning on FPGA-based SoCs. Specifically, we analyze the various stages of the Epsilon-Greedy algorithm for RL and propose a novel microarchitecture that reduces the latency by optimizing the memory access during each iteration. Consequently, we present multiple designs that provide varying trade-offs between performance, power dissipation, and resource utilization of the accelerator. With the proposed approach, we report considerable improvement in throughput with lower resource utilization over state-of-the-art design implementations." @default.
- W3175606556 created "2021-07-05" @default.
- W3175606556 creator A5002474306 @default.
- W3175606556 creator A5028792027 @default.
- W3175606556 creator A5051064456 @default.
- W3175606556 creator A5076463582 @default.
- W3175606556 date "2021-06-22" @default.
- W3175606556 modified "2023-10-16" @default.
- W3175606556 title "<i>MemOReL</i>" @default.
- W3175606556 cites W2113226229 @default.
- W3175606556 cites W2145339207 @default.
- W3175606556 cites W2471791811 @default.
- W3175606556 cites W2771384842 @default.
- W3175606556 cites W2901621510 @default.
- W3175606556 cites W2904195769 @default.
- W3175606556 cites W2915058268 @default.
- W3175606556 cites W2938157874 @default.
- W3175606556 cites W2998165854 @default.
- W3175606556 cites W3089396352 @default.
- W3175606556 doi "https://doi.org/10.1145/3453688.3461533" @default.
- W3175606556 hasPublicationYear "2021" @default.
- W3175606556 type Work @default.
- W3175606556 sameAs 3175606556 @default.
- W3175606556 citedByCount "1" @default.
- W3175606556 countsByYear W31756065562022 @default.
- W3175606556 crossrefType "proceedings-article" @default.
- W3175606556 hasAuthorship W3175606556A5002474306 @default.
- W3175606556 hasAuthorship W3175606556A5028792027 @default.
- W3175606556 hasAuthorship W3175606556A5051064456 @default.
- W3175606556 hasAuthorship W3175606556A5076463582 @default.
- W3175606556 hasConcept C108583219 @default.
- W3175606556 hasConcept C111919701 @default.
- W3175606556 hasConcept C113775141 @default.
- W3175606556 hasConcept C115903868 @default.
- W3175606556 hasConcept C118524514 @default.
- W3175606556 hasConcept C120314980 @default.
- W3175606556 hasConcept C126255220 @default.
- W3175606556 hasConcept C149635348 @default.
- W3175606556 hasConcept C154945302 @default.
- W3175606556 hasConcept C157764524 @default.
- W3175606556 hasConcept C173608175 @default.
- W3175606556 hasConcept C206729178 @default.
- W3175606556 hasConcept C26713055 @default.
- W3175606556 hasConcept C33923547 @default.
- W3175606556 hasConcept C41008148 @default.
- W3175606556 hasConcept C42935608 @default.
- W3175606556 hasConcept C555944384 @default.
- W3175606556 hasConcept C76155785 @default.
- W3175606556 hasConcept C82876162 @default.
- W3175606556 hasConcept C97541855 @default.
- W3175606556 hasConceptScore W3175606556C108583219 @default.
- W3175606556 hasConceptScore W3175606556C111919701 @default.
- W3175606556 hasConceptScore W3175606556C113775141 @default.
- W3175606556 hasConceptScore W3175606556C115903868 @default.
- W3175606556 hasConceptScore W3175606556C118524514 @default.
- W3175606556 hasConceptScore W3175606556C120314980 @default.
- W3175606556 hasConceptScore W3175606556C126255220 @default.
- W3175606556 hasConceptScore W3175606556C149635348 @default.
- W3175606556 hasConceptScore W3175606556C154945302 @default.
- W3175606556 hasConceptScore W3175606556C157764524 @default.
- W3175606556 hasConceptScore W3175606556C173608175 @default.
- W3175606556 hasConceptScore W3175606556C206729178 @default.
- W3175606556 hasConceptScore W3175606556C26713055 @default.
- W3175606556 hasConceptScore W3175606556C33923547 @default.
- W3175606556 hasConceptScore W3175606556C41008148 @default.
- W3175606556 hasConceptScore W3175606556C42935608 @default.
- W3175606556 hasConceptScore W3175606556C555944384 @default.
- W3175606556 hasConceptScore W3175606556C76155785 @default.
- W3175606556 hasConceptScore W3175606556C82876162 @default.
- W3175606556 hasConceptScore W3175606556C97541855 @default.
- W3175606556 hasLocation W31756065561 @default.
- W3175606556 hasOpenAccess W3175606556 @default.
- W3175606556 hasPrimaryLocation W31756065561 @default.
- W3175606556 hasRelatedWork W2039520903 @default.
- W3175606556 hasRelatedWork W2064530646 @default.
- W3175606556 hasRelatedWork W2096823099 @default.
- W3175606556 hasRelatedWork W2163403354 @default.
- W3175606556 hasRelatedWork W2400714260 @default.
- W3175606556 hasRelatedWork W2524802307 @default.
- W3175606556 hasRelatedWork W2547383453 @default.
- W3175606556 hasRelatedWork W2949525946 @default.
- W3175606556 hasRelatedWork W4293143575 @default.
- W3175606556 hasRelatedWork W2506672464 @default.
- W3175606556 isParatext "false" @default.
- W3175606556 isRetracted "false" @default.
- W3175606556 magId "3175606556" @default.
- W3175606556 workType "article" @default.