DataTau
new
|
comments
|
leaders
|
submit
login
1 point
by
aqokaq
3128 days ago |
link
|
parent
Deep-Q Learning algorithm can be made a bit better by using separate functions for selection and evaluation of actions:
Deep Reinforcement Learning with Double-Q Learning
http://arxiv.org/pdf/1509.06461v1.pdf
RSS
| Announcements