Deep Q Learning for Video Games - The Math of Intelligence #9

  • 4 Replies
  • 342 Views
*

keghn

  • Trusty Member
  • ********
  • Replicant
  • *
  • 603

Deep Q Learning for Video Games - The Math of Intelligence #9 : 


*

keghn

  • Trusty Member
  • ********
  • Replicant
  • *
  • 603
Re: Deep Q Learning for Video Games - The Math of Intelligence #9
« Reply #1 on: August 13, 2017, 05:57:48 pm »
Watch Elon Musk's OpenAI DOTA 2 Bot Destroy The World’s Best Players At The International: 

https://hothardware.com/news/watch-elon-musks-openai-dota-2-bot-destroy-the-worlds-best-players-at-the-international

*

yotamarker

  • Trusty Member
  • ********
  • Replicant
  • *
  • 588
Re: Deep Q Learning for Video Games - The Math of Intelligence #9
« Reply #2 on: August 13, 2017, 09:21:20 pm »
if as he said long time reward = less importance
wouldn't it mean Mario won't search tough hidden bosses and bonus stages ?

*

LOCKSUIT

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1125
  • First it wiggles, then it is rewarded.
    • Enter Lair
Re: Deep Q Learning for Video Games - The Math of Intelligence #9
« Reply #3 on: August 13, 2017, 11:36:05 pm »
But at the end of Mario, he still never gets Peach.

Get that.

Oh the hunt. The man has great desire let me tell ya.

*

keghn

  • Trusty Member
  • ********
  • Replicant
  • *
  • 603
Re: Deep Q Learning for Video Games - The Math of Intelligence #9
« Reply #4 on: September 14, 2017, 02:14:57 pm »

A Guide to DeepMind's StarCraft AI Environment: