Intelligent Trainer for Dyna-Style Model-Based Deep Reinforcement Learning.
IEEE Trans Neural Netw Learn Syst. 2021 Jun;32(6):2758-2771. doi: 10.1109/TNNLS.2020.3008249. Epub 2021 Jun 2.
IEEE Trans Neural Netw Learn Syst. 2021.
PMID: 32866102
This decomposition approach provides much-needed flexibility to implement different trainer designs, referred to "train the trainer." In our research, we propose and optimize two alternative trainer designs: 1) an unihead trainer and 2) a multihead …
This decomposition approach provides much-needed flexibility to implement different trainer designs, referred to "train the traine …