Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Reinforcement Learning: An Introduction (2018) [pdf]

incompleteideas.net

Post reply on HN
Print view View original

1–10 of 26 posts

Re: Reinforcement Learning: An Introduction (2018) [pdf]

#2

Post by keeptrying »

The authors , Barto and Sutton take such a complicated subject and explain it in such simple prose.

I don’t think think I’ve read any other work that does this as well.

Also RL is only going to grow in use and popularity. Highly recommend it for mL practitioners.

Re: Reinforcement Learning: An Introduction (2018) [pdf]

#4

Post by abhgh »

keeptrying wrote: post #2

The authors , Barto and Sutton take such a complicated subject and explain it in such simple prose. I don’t think think I’ve read any other work that does this as well. Also RL is only going to grow in use and popularity. Highly recommend it for mL practitioners.

I hope it grows in popularity if only because its an interesting take on learning. I did a course on RL in 2007 and our textbook was the 1st edition of this book - back then, it was perceived to be a very niche area and a lot of ML practitioners (there weren't many of those either :) ) had only just about heard of RL.

I am happy that it's popular today.

Re: Reinforcement Learning: An Introduction (2018) [pdf]

#6

Post by nafizh »

With all its hype in RL, I am yet to see significant real life problems solved with it. I am afraid with all the funding going into it, and nothing to show for except being able to play complex games, this might contribute to the mistrust in proper utilization of research funds. Also the reproducibility problem in RL is many times worse than in ML.

Re: Reinforcement Learning: An Introduction (2018) [pdf]

#7

Post by platz »

abhgh wrote: post #4
keeptrying wrote: post #2

The authors , Barto and Sutton take such a complicated subject and explain it in such simple prose. I don’t think think I’ve read any other work that does this as well. Also RL is only going to grow in use and popularity. Highly recommend it for mL practitioners.

I hope it grows in popularity if only because its an interesting take on learning. I did a course on RL in 2007 and our textbook was the 1st edition of this book - back then, it was perceived to be a very niche area and a lot of ML practitioners (there weren't many of those either :) ) had only just about heard of RL. I am happy that it's popular today.

> I am happy that it's popular today.

For doing what?

Re: Reinforcement Learning: An Introduction (2018) [pdf]

#8

Post by siekmanj »

platz wrote: post #7
abhgh wrote: post #4

Earlier quoted context omitted.

I hope it grows in popularity if only because its an interesting take on learning. I did a course on RL in 2007 and our textbook was the 1st edition of this book - back then, it was perceived to be a very niche area and a lot of ML practitioners (there weren't many of those either :) ) had only just about heard of RL. I am happy that it's popular today.

> I am happy that it's popular today. For doing what?

It's definitely finding a niche in robotic control. My lab just released a paper about running a policy trained in simulation in the real world on a bipedal robot.

Re: Reinforcement Learning: An Introduction (2018) [pdf]

#9

Post by imh »

nafizh wrote: post #6

With all its hype in RL, I am yet to see significant real life problems solved with it. I am afraid with all the funding going into it, and nothing to show for except being able to play complex games, this might contribute to the mistrust in proper utilization of research funds. Also the reproducibility problem in RL is many times worse than in ML.

I'd bet that sample efficiency is a factor in translating they most hyped bits of RL into solving IRL problems. So many business problems translate to "Learn which of these things to do, as quickly and cheaply as possible."

Re: Reinforcement Learning: An Introduction (2018) [pdf]

#10

Post by svalorzen »

If you ever feel like trying out the algorithms contained in the book without going to the trouble of reimplementing everything from scratch feel free to come over to https://github.com/Svalorzen/AI-Toolbox. This is a library I have maintained during the past 5 years and implements quite a lot of RL algorithms, and can be used with both C++ and Python. It's very focused on being understandable and having a clear documentation, so I'd love to help you out starting up :)
Post reply on HN