Re:Human-level control through deep reinforcement learning
by fchollet on Wednesday February 25, 2015 @05:45PM
(#9110303)
Interesting that they're using Torch7. The code is pretty concise and readable, very cool stuff.
Re:Human-level control through deep reinforcement learning
by yablak on Wednesday February 25, 2015 @06:26PM
(#9110474)
They wrote Torch7...
Re:Human-level control through deep reinforcement learning
by fchollet on Wednesday February 25, 2015 @07:38PM
(#9110714)
I'm sure a number of people who have contributed to Torch are working at DeepMind. However, Torch has been around for much longer than DeepMind (about 12 years at this point). Two of the major contributors to Torch, Ronan Collobert and Clement Farabet, were never DeepMind employees.
To be fair, another major contributor to Torch is a co-author of this paper (Kavukcuoglu).
If you're interested, one of the main authors (David Silver) teaches a very good and intuitive introductory class on reinforcement learning at UCL: http://www0.cs.ucl.ac.uk/staff/d.silver/web/Teaching.html