Human-level control through deep reinforcement learning : Nature : Nature Publishing Group http://t.co/lfYLdmAMFV
このところGoogleのDQN http://t.co/EWylhjXZSU の論文とコードを読んでたのだけど、やっと大体理解できた。特にコードがLUA + torchで書かれてたのだけど、ものすごく読みにくい。 あと、結果には問題ないけど、バグも見つけた。
http://t.co/kgdU4SX0Pb Human-level control through deep reinforcement learning
deep RL: http://t.co/F142dRAsTs
What happens when AI techniques of deep learning + reinforcement lrng are combined? Atari mastery. http://t.co/kTferDrXGM #2MA
What happens when AI techniques of deep learning + reinforcement lrng are combined? Atari mastery. http://t.co/kTferDrXGM #2MA
What happens when AI techniques of deep learning + reinforcement lrng are combined? Atari mastery. http://t.co/kTferDrXGM #2MA
Human-level control through deep reinforcement learning : Nature http://t.co/NBNRayxDbb
Human-level control through deep reinforcement learning http://t.co/zeordGc3lH
#svpol Sakta men säkert närmar vi oss singulariteten. http://t.co/2MjhMmwip8
The research on the Atari game-mastering AI system is now out. Prepare to be amazed: http://t.co/kTferDrXGM #2MA
The research on the Atari game-mastering AI system is now out. Prepare to be amazed: http://t.co/kTferDrXGM #2MA
The research on the Atari game-mastering AI system is now out. Prepare to be amazed: http://t.co/kTferDrXGM #2MA
@tmaehara どっちかっていうとこれですね http://t.co/0vhZUcyclC
MT @calcworks Human-level control through deep #reinforcementlearning: http://t.co/fnaIhwmEPu #deeplearning #machinelearning
MT @calcworks Human-level control through deep #reinforcementlearning: http://t.co/fnaIhwmEPu #deeplearning #machinelearning
MT @calcworks Human-level control through deep #reinforcementlearning: http://t.co/fnaIhwmEPu #deeplearning #machinelearning
@catrobotil @amirvaxman כמו כן, קורה שמאמר מדעי המחשב או דומה לו מסתנן לשם: http://t.co/1QrlRjGuCe ואפילו מהתחום שלי בדוקטורט היה אחד.(נדיר)
http://t.co/apdxDvPSGp artificial neurons learned to play Atari video games better than humans #AI #VideoGame @atari @NatureNews
@stanfordnlp @TariqDaouda What about these? http://t.co/gjWx2fWJFZ http://t.co/UlVbr8NkYB
Deep learning AI program plays Space Invaders! (Video) http://t.co/LWJX28Ollk
algorithm learns to play atari better than humans in a few hours, with only pixels and score as input http://t.co/MbYIYRuvsM
@caiocgomes Te mais esse tb: http://t.co/sipnoI2bDj ;-)
Google DeepMind paper on Nature. Full algorithm is available as well! http://t.co/NshuDhy7XL Such a cool idea. I'm so excited for the future
A simulated network of artificial neurons learned to play Atari video games better than humans in a matter of hours http://t.co/8IIxf0QOh9
Artificial neural networks by Google DeepMind outplay humans on Atari games with only pixels and score as inputs. http://t.co/37Rm6qIakS
Artificial neural networks by Google DeepMind outplay humans on Atari games with only pixels and score as inputs. http://t.co/37Rm6qIakS
Had not seen this earlier. Really impressive combination of reinforcement learning and deep NN: http://t.co/PjmqxWqkxV
Had not seen this earlier. Really impressive combination of reinforcement learning and deep NN: http://t.co/PjmqxWqkxV
http://t.co/Zz4nJbXGUS Human-level control through deep reinforcement learning
Human-level control through deep reinforcement learning: http://t.co/ckTEZlJbGB #deeplearning #machinelearning #reinforcementlearning
Human-level control through deep reinforcement learning: http://t.co/ckTEZlJbGB
Surely you've heard from them if you're into AI and General Video Game Playing. Their paper in Nature, DNN for GVGP: http://t.co/j554j0MF4w
Human-level control through deep reinforcement learning http://t.co/4U39zTeSMS #AI #machinelearning #ML #deeplearning #RL
Human-level control through deep reinforcement learning http://t.co/4U39zTeSMS #AI #machinelearning #ML #deeplearning #RL
Human-level control through deep reinforcement learning http://t.co/4U39zTeSMS #AI #machinelearning #ML #deeplearning #RL
Human-level control through deep reinforcement learning http://t.co/4U39zTeSMS #AI #machinelearning #ML #deeplearning #RL
Human-level control through deep reinforcement learning http://t.co/TLIqlbEmp5
Algorithm that genuinely learns 2 learn & unlearn - not AI but an "artificial-artificial intelligence" breakthrough http://t.co/NJVDur8sT6
@h0pbeat They seem to suck at Montezuma's Revenge http://t.co/8k75vB9lLf (fig 3) http://t.co/xg59VfvJGn
Never thought knowledge of Ms. Pacman would come in handy in a journal club discussion. http://t.co/4TRy2yQ2IK #deepmind #deeplearning
They are coming. http://t.co/Cb8gKGcQzF
http://t.co/gx38QQxPys Human-level control through deep reinforcement learning
Human-level control through deep reinforcement learning http://t.co/FyrlIfXv1n
"Human-level control through deep reinforcement learning" new Nature paper from DeepMind! http://t.co/WDefOyiWZ3
Human-level control through deep reinforcement learning. http://t.co/tttEBn6Yo5
deep Q-networkか。まだしっかり読んでないけどQって何なん?QuestionのQ? Human-level control through deep reinforcement learning http://t.co/JDxV57rvbO
Very interesting research at DeepMind google's lab in London https://t.co/KISaXtUtKE
Watch neural nets learn to play computer games: Human-level control through deep reinforcement learning via @nature http://t.co/JOjlFwSJfR
@oobr Glad it's not just me... :) And I'd thought pictures of Space Invaders in Nature was good enough...! http://t.co/59VkwBpTU5
Raia Hadsell: Google DeepMind gets Nature cover article for reinforcement learning to play Atari games http://t.co/KRg3a7hWe9 #WTM15
Raia Hadsell: Google DeepMind gets Nature cover article for reinforcement learning to play Atari games http://t.co/KRg3a7hWe9 #WTM15
【情報科学】【人工知能】グーグル傘下のDeepMind社が、画期的な人工知能アルゴリズムを開発し、ネイチャーで発表した。自力で試行錯誤し、ブロック崩しなどのゲームで、プロの人間を凌駕。ただし、名前は「DQN」。日本ではウケなそう。 http://t.co/UFb7HrQ1ug
"We set out to create a single algorithm..." - Human-level control through deep reinforcement learning http://t.co/wRAZKA8I8P
Demonstration of Learning Progress in da Game Breakout Human-level control through deep reinforcement learning Nature http://t.co/ba1h5tqIvV
The same neural network training algorithm learned to play 49 Atari 2600 games, solely from gameplay pixels http://t.co/ebZBotxMYD
Is this #EthicalAI? Hard to know! "Deep Reinforcement Learning Our alumni" http://t.co/Exf23CENIS #deeplearning
@usethespacebar I got you http://t.co/AzItmideIP
Human-level control through deep reinforcement learning http://t.co/mTW2EzKpqP
A computer algorithm that can play Atari games... and is really good at them! #skynet http://t.co/mmpACHtuBt
Human-level control through deep reinforcement learning #AI - check out video 2 http://t.co/qQplEUEKta
An artificial agent can learn to play 49 classic arcade video games when given only minimal background information http://t.co/WP500GNht9
Peter Norvig at #cscw2015 showing off Googlers teaching a computer to play Breakout based on images & score alone. http://t.co/zazrVwWkBy
@Google's Norvig on teaching a computer to become a Breakout-hero through image recognition http://t.co/rlwjl2Gi3K #cscw2015 #las2015
Peter Norvig at #cscw2015 showing off Googlers teaching a computer to play Breakout based on images & score alone. http://t.co/zazrVwWkBy
"Human-level control through deep reinforcement learning". http://t.co/7PQYeopFQk. Peter Norvig, #2015las, #cscw2015.
http://t.co/T1kOxhP9vU Nature vs. Machine learning #AI
原論文は V. Mnih et al., Nature 518, 529 (2015) http://t.co/XBpLtEkRw8 ちなみにここでも dimensionality が俎上に。。。
深層学習がnatureに載る時代ですね.http://t.co/HgbYihAdoQ
深層学習がnatureに載る時代ですね.http://t.co/HgbYihAdoQ
Light weekend reading in case you want to geek out. Nature paper explaining how #deeplearning used to play atari. http://t.co/OBDITFmcR9
Light weekend reading in case you want to geek out. Nature paper explaining how #deeplearning used to play atari. http://t.co/OBDITFmcR9
"Real stupidity beats artificial intelligence every time" #PratchettQuotes http://t.co/BcJzWeQnfK esp. on Ms. Pac Man http://t.co/gqR0SsdUOW
Human-level control through deep reinforcement learning #AI http://t.co/jA7ScGbtED
Hope for Python school Pelita programmers: Humans still better at Pacman. http://t.co/hD1vymGAFF (via @MInkorrekt) // http://t.co/KcBjTeaYfk
Experience replay (#hippocampus) turns out to be useful... for artificial reinforcement learning. http://t.co/u2587G3HsT.
Human-level control through deep reinforcement learning http://t.co/07riGy3duO
http://t.co/cWpYqrN0Gb Human-level control through deep reinforcement learning
A landmark paper in AI - deep reinforcement learning - a general intelligence. From Demis Hassabis and co: http://t.co/ksjfUq1g4s
Human-level control through deep reinforcement learning http://t.co/uaiqWMTdVB (how a computer can train itself to beat us on classic games)
Nature paper "Human-level control through deep reinforcement learning" can be read here http://t.co/Pwfc2PLZ9k
Experience replay (#hippocampus) turns out to be useful... for artificial reinforcement learning. http://t.co/u2587G3HsT.
Google Deep Mind AI learns to outperform humans playing vintage arcade games http://t.co/2e0zpK4fAz
What happens when AI techniques of deep learning + reinforcement lrng are combined? Atari mastery. http://t.co/kTferDrXGM #2MA
Human-level control through deep reinforcement learning #OneToOneLearning http://t.co/wPnyR5R5Zb
Human-level control through deep reinforcement learning http://t.co/4qLsaDZGIw Inside DeepMind: http://t.co/hXWPGGDYDY #science
Human-level control through deep reinforcement learning http://t.co/4qLsaDZGIw Inside DeepMind: http://t.co/hXWPGGDYDY #science
An artificial agent can learn to play 49 classic arcade video games when given only minimal background information http://t.co/WP500GNht9
Unlike other uses of deep learning limited to perception, deepmind integrates it in an agent that _acts_ on the world http://t.co/c9MWzY1JSh
An artificial agent can learn to play 49 classic arcade video games when given only minimal background information http://t.co/WP500GNht9
Deep reinforcing learning with only human applicable inputs( pixels, game scores) applied to breakout over time http://t.co/hMmeVjHpDD
Human-level control through deep reinforcement learning : on Nature : @NatureNews http://t.co/phyZImQAmS
自力で学習する人工知能「DQN」、ゲーム攻略で習熟力披露 AFP http://t.co/igWaYm74Tl ルールも説明書も与えられない状態からdeep Q-networkはAtari2600で人間以上に上達する(nature)→ http://t.co/lBQM9HQ5cX
#deeplearning algorithms http://t.co/8uVmtKshFe (they can be really good at playing Atari2600 games)
#deeplearning algorithms http://t.co/tQ72ml2JRt (they can be really good at playing Atari2600 games)
#deepmind http://t.co/Gw5kgQuTdy great research and an awesome paper #deeplearning That's the future.
Human-level control through deep reinforcement learning--Google's #DeepMind paper w/ Q-Network algorithm http://t.co/YBxsEuk0s6 #atari2600