“If you want to see the state of the art in using AI to play video games, look no further than "Playing Atari with Deep Reinforcement Learning" [2], where a single general AI system learns to play many different games.”
Playing Atari with Deep Reinforcement Learning
卷积网络加 Q 学习,从原始像素学会玩 Atari
提出首个能直接从高维感知输入成功学习控制策略的深度学习模型:用 Q-learning 变体训练卷积神经网络,输入是原始像素,输出是估计未来回报的价值函数。在 Arcade Learning Environment 的七款 Atari 游戏上,同一架构和算法不作调整,在六款上超过此前所有方法,其中三款超过人类专家。
评论者把它当作 DeepMind 深度强化学习的起点:同一个系统只看像素和得分信号就能学会多款游戏,是 AlphaGo 之前 AI 玩游戏的代表性进展。
HN 书评
推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及“the readme links a cool paper on applying convolutional neural networks to old games.”
“While this complaint generally has validity, their paper [1] does IMO present an advance; it's not just handing a bunch of labeled data off to a large neural network.”
“There is a paper on Arxiv that describes the approach”
“Also, they have published ArXiv paper on the subject: Playing Atari with Deep Reinforcement Learning.”
“which is supposedly just an algorithm working on the raw pixels and then beating all other programs and half of the human experts.[1]”
看全部 16 条 →
“A note -- if you're linking to arXiv, it's better to link to the abstract rather than directly to the PDF.”
“Here's the original research paper if you're interested.”
“DeepMind has a paper describing this approach: Playing Atari with Deep Reinforcement Learning”
“The score was separate, or rather a signal representing increased score.”
“they use four frames in sequence. That was at the end of 2013.”
“Deepmind has been focussing on this since the beginning[0] and actually has great progress as far as Atari games go[1].”
“Before AlphaGo, DeepMind released a reinforcement learning algorithm that could play many Atari games just from the raw pixels on the screen, in many games surpassing humans.”
“They don't mention this in the paper, but the network wasn't trained to convergence at each training step”
“Third paragraph of Intro in original DeepMind paper Playing Atari with Deep Reinforcement Learning”
“some of DeepMind's earliest work was a deep reinforcement learning model that could play seven different Atari games with no adjustments needed to switch games[2]”
- Playing Atari with Deep Reinforcement Learning ↗bentoner · 2014-01-13 · 3 分
- Deep Mind's neural networks play 7 Atari 2600 games with more skill than a human ↗JumpCrisscross · 2014-01-28 · 5 分
- Playing Atari with Deep Reinforcement Learning [pdf] ↗bradneuberg · 2014-10-20 · 37 分
- Playing Atari with Deep Reinforcement Learning (2013) [pdf] ↗jeremynixon · 2015-01-03 · 21 分