Comparison of DQN and its variants.
- Specify hyper-parameters in
config.py
. - Run
./run.sh $device $algorithm
for training with three random seeds. - Visualize the results by
python plot.py *log*
.
@book{deepRL-2020,
title={Deep Reinforcement Learning: Fundamentals, Research, and Applications},
editor={Hao Dong, Zihan Ding, Shanghang Zhang},
author={Hao Dong, Zihan Ding, Shanghang Zhang, Hang Yuan, Hongming Zhang, Jingqing Zhang, Yanhua Huang, Tianyang Yu, Huaqing Zhang, Ruitong Huang},
publisher={Springer Nature},
note={\url{http://www.deepreinforcementlearningbook.org}},
year={2020}
}