Hello Rohan, the timeline looks reasonable to me, I guess you might need more time for the ACER implementation, just because this is the first implementation.
Thanks, Marcus > On 24. Mar 2018, at 19:43, Rohan Raj <[email protected]> wrote: > > Dear Marcus, > > I am Rohan Raj,aka luffy1996 on github and IRC channel. I am currently > drafting my gsoc proposal for mlpack in reinforcement learning project . I > would like to know your views on this proposal. > > I am planning to go ahead with PPO , ACER and ACKTR implementation on > mlpack for GSOC 2018 . I would also like to mention that I am planning to > give 3 - 3.5 weeks for each of these algorithms. This makes my tentative > schedule as > Week 1 : Implementation of continuous action space games for Actor critic > Algorithms. > Week 2 - 4 : ACER > Week 5 - 7 : A2C(synchronous A3C) and PPO > Week 8 - 10 : ACKTR > Week 11 - 12 : Bug fixing and final submission. > > It would be grate if you can provide your input to this tentative schedule. I > would be happy to add changes based on your reviews and suggestions. > > PPO : (https://arxiv.org/abs/1707.06347 <https://arxiv.org/abs/1707.06347>) > ACER : (https://arxiv.org/abs/1611.01224 <https://arxiv.org/abs/1611.01224>) > ACKTR : (https://arxiv.org/abs/1708.05144 <https://arxiv.org/abs/1708.05144>) > > Rohan Raj > Indian Institute of Technology Guwahati > Assam , India > Phone : +91 8723990557 > > ᐧ
_______________________________________________ mlpack mailing list [email protected] http://knife.lugatgt.org/cgi-bin/mailman/listinfo/mlpack
