maybedy / RLStudy

Team study repository

Reinforcement Learning Study

• Team study repository
• Member : kim, yang, moon, jang, oh, yu, yu, seok

Material

Arthur Juliani's "Simple Reinforcement Learning with Tensorflow" post series

• 20170104 ~ 20170308
• See here(Part 0 of the series) -> https://medium.com/emergent-future/simple-reinforcement-learning-with-tensorflow-part-0-q-learning-with-tables-and-neural-networks-d195264329d0

Rich. Sutton's reinforcement learning book(2nd edition, 2017 draft)

• 20170315 ~ http://incompleteideas.net/sutton/book/the-book-2nd.html

Weekly plan

Aw Juliani's post series

• Part 0 : Oh
• Part 1 : Moon
• Part 1.5 : Jang
• Part 2 : Yang
• Part 3 : Kim
• Part 4 : Yu(MOON)
• Part 5 : Yu(SH)
• Part 6 : Oh -> pending : thursday(02/23)
• Part 7 : Moon
• Part 8 : Jang

Rich. Suttion's RL book

• RL Ch 3 : Yang
• RL Ch 4 : Kim
• RL Ch 5 : Yu(MOON)
• RL Ch 6 : Yu(SH)
• RL Ch 7 : Oh
• RL Ch 8 : Moon
• RL Ch 9 : Jang
• RL Ch 10 : Seok
• RL Ch 11 : Yang
• RL Ch 12 : Kim
• RL Ch 13 : Yu(MOON)
• RL Ch 16 : Moon
• ... : Oh
• ... : Yu(SH)
• ... : Jang
• ... : Seok
• ... : Yang
• ... : Kim

Candidates

• 1. NIPS 2017 :
  • It is hard to find papers to study
• 2. Reinforcement learning papers :
• 3. Deep RL papers : better archiecture or high performance

• Conclusion : NIPS or DeepMind or Well-made papers

ETC

TODO

• Make some helpful material list in 'Readme' or other markdown file
• Add course materials for each part

References

• http://selfdrivingcars.mit.edu/
• https://hunkim.github.io/ml/
• http://aiworld.io/
• https://github.com/maybedy/RLStudy/blob/master/reading_list.md
• http://www0.cs.ucl.ac.uk/staff/d.silver/web/Teaching.html