Introduction to Reinforcement Learning (2015)

Re:Introduction to Reinforcement Learning (2015) by bitL on Sunday December 27, 2020 @08:26AM (#25550867)

RL needs a supercomputer and its code is usually too fragile - making a trivial mistake anywhere (missing a constant multiplication, swapping the order of two consecutive lines of code etc.) would likely lead to your model never converging even if you got everything else right.

[ Parent | on HN ]

Re:Introduction to Reinforcement Learning (2015) by chasely on Sunday December 27, 2020 @09:28AM (#25551245)

The hard part of RL for the problems I've encountered in my work is that you need a simulator. Building a reliable and accurate simulator is often an immense undertaking.

[ Parent | on HN ]

Re:Introduction to Reinforcement Learning (2015) by dgb23 on Sunday December 27, 2020 @10:21AM (#25551603)

Maybe data scientists should team up (more?) with game programmers. They have a ton of experience in building very complex simulations.

[ Parent | on HN ]

Re:Introduction to Reinforcement Learning (2015) by Ma8ee on Sunday December 27, 2020 @08:31AM (#25550898)

Which code is not fragile in that sense? I think that is a rather strange criticism.

[ Parent | on HN ]

Re:Introduction to Reinforcement Learning (2015) by Iv on Sunday December 27, 2020 @08:34AM (#25550915)

You can do RL on an raspberry pi. Depends what problem you are trying to solve but not all of them require video analysis and billions of parameters.

[ Parent | on HN ]

Re:Introduction to Reinforcement Learning (2015) by cbames89 on Sunday December 27, 2020 @08:43AM (#25550968)

Technical point: Value functions that are a constant multiples of each other result in the same behavior.

[ Parent | on HN ]

Re:Introduction to Reinforcement Learning (2015) by bitL on Monday December 28, 2020 @10:03AM (#25559679)

Making a constant multiplication mistake somewhere in the code doesn't imply the new value function would be a constant multiply of the optimal one.

[ Parent | on HN ]