News for hackers. Stuff that matters.

adamheins

Show HN: Play tag against an opponent trained with reinforcement learning

https://adamheins.com/projects/shadows/web/

This is a little prototype I made recently while learning about reinforcement learning and to explore the use of machine learning in games. The learned model that the computer-controlled opponent uses is by no means optimal, but I think it still gives an interesting experience. Enjoy!

( Read More... | no comments | on HN )
Re:Ask HN: Books you read in 2017? on Tuesday December 19, 2017 @11:36AM (#15961393)
To continue the Ishiguro and Hemingway theme, two of the books I enjoyed most this year were The Buried Giant (Ishiguro) and The Old Man and the Sea (Hemingway).
Re:Humble People Make the Best Leaders on Thursday April 13, 2017 @02:21PM (#14109577)
Your description of the Chinese system reminds me of the cursus honorum[1] from Ancient Rome.

[1] https://en.wikipedia.org/wiki/Cursus_honorum

Re:Voice Calls: Secure, Crystal-Clear, AI-Powered on Thursday March 30, 2017 @12:25PM (#13996748)
I don't know much about the different router brands and models. Which would you consider to be some of the good ones?
Re:The San Francisco Coffee Crawl on Friday December 30, 2016 @08:38PM (#13288923)
What's the coffee scene like in Warsaw and Poland in general?

Also, switching to decaf partway through is probably the smart thing to do :)