赞
踩
最近研究AI理论,发现一篇Marcus Hutter关于入门学习AI技术提出的建议的文章。感觉很有启发,特翻译成中文与诸位童鞋共同分享。为了避免翻译时引入的歧义、错误,特别保留了英文原文,供大家对比着查阅。谢谢。
随着人工智能在我们生活中的影响不断增大,给我们的生活带来了越来越多的便利。很多同学开始喜欢并想进入“人工智能”领域一试身手。以下是澳大利亚国立大学的Marcus Hutter关于入门学习人工智能的一些建议。
一、关于通用人工智能模型(AIXI)
首先,要了解AI基础知识和建模的一般性原理及建议,请参见《通用人工智能——基于概率算法的序贯决策》(Universal Artificial Intelligence——Sequential Decisions based on Algorithmic Probability)【国内暂时没这书 -_-!】;
其次,基于“序贯决策”(Sequential decision theory)和“索洛莫诺夫通用(一般性)归纳”(Solomonoff’s theory of universal induction)理论,澳大利亚国立大学的Marcus Hutter提出了AIXI模型,该模型的具体知识,请参见Marcus Hutter的论文《UNIVERSAL ALGORITHMIC INTELLIGENCE A mathematical top→down approach》;
事实证明,目前的技术水平下AIXI模型的理论意义大于实际技术实现意义。因为AIXI模型的计算复杂度是δ(4)。就是说不仅这个模型是不可计算的,而且它的不可计算等级为4(一般的停机问题为δ(1))!也即是说目前只能采用可计算的模型来尽可能趋近AIXI模型的效果,目前热门的deepmind,就是这个方法的先行者(deepmind的创始人是marcus hutter的博士)。
虽然AIXI模型目前暂时无法实际实现,但是作为理论学习的指导是很有意义的,它通过这一个模型(公式),就可以概括深度学习、SVM、传统逻辑、强化学习等各种理论。
AIXI模型的公式如下:
如果你对下文中“更多参考资料”的数量、种类、内容复杂度等方面有困惑,我建议你可以先翻阅Richard S. Sutton和 Andrew G. Barto编著的《强化学习》(Reinforcement Learning)。
在学习它之前不需要具备什么背景知识。该书深入浅出的描述了一些关键思想,开放性问题,以及这个领域的大量应用示例。不要对这本“如此简单”的书感到惊讶,它主要讲解AI相关的基本知识和概念,而不是证明其在实际实现时的正确性。如果它里面详细讲解在实践中如何实现功能或者证明其理论的正确性,这会使本书非常晦涩难懂和不易学习。
学习了《强化学习》,对“人工智能”概念有了一定感性认识后就可以开始翻阅另一本由Stuart Russell和Peter Norvig编写的人工智能专著《人工智能——一种现代的方法(第3版)》(去x宝购物平台上随便一搜就可以买到),它为“通用人工智能”提供了更为全面的讲解和概括。最后,当你对“人工智能”有了基本了解并且具备一定理论知识基础后,可以进一步学习由Li和Vitanyi 编著的关于算法信息论论文《柯尔莫戈洛夫复杂性》,它对“算法信息论”进行了极好的论述和证明。
当你通过上述学习已经具备一些基本的“决策理论”和“算法信息论”背景知识后,那么你或许会对进一步深入研究“通用人工智能(Artificial General Intelligence/Universal Artificial Intelligence,AGI/UAI)”理论产生浓厚兴趣。
二、更多参考资料
写给急于开始“创建智能机器”的人:
如果你是一个急切的想马上建立超级智能机器,而又不想浪费太多时间阅读或学习的学生,那么好吧,我可以坦率的告诉你:其他人在过去的50年里也这样尝试过,失败了,你也不会例外。如果你依然坚持这种做法,那么至少要好好读一下Shane Legg撰写的《Machine Super Intelligence》一文。这是一篇优秀的非技术性论文,论述了超级智能机器的必要组成部分。但它并不会绕过你在一般性人工智能理论上缺失知识的问题,而帮你快速建立一个“缩减版”的实际存在的人工智能算法。然而,它可能会促使你考虑阅读我现在为你推荐的书籍。
For the impatient. If you are the sort of impatient student who wants to build super intelligent machines right away without "wasting" time reading or learning too much, well, others have tried in the last 50 years and failed, and so will you. If you can't hold back, at least read Legg (2008) [Leg08]. This is an excellently written non-technical thesis on the necessary ingredients for super intelligent machines.
It will not help you much building one, since in order to properly understand the general theory and to bridge the gap to "narrow" but practical existing AI algorithms, you need a lot more background. Nevertheless, [Leg08] might motivate you to consider reading the books I'll recommend now.
关于人工智能:
Stuart Russell和Peter Norvig编写的《人工智能——一种现代的方法(第3版)》(去x宝购物平台上随便一搜就可以买到),是一本学习人工智能的课本。这本书提供了一个广泛的可靠的关于人工智能全方位背景的内容介绍和概观。该书在人工智能学习过程中的作用可以说是“无可替代”的。不管您以后专门从事人工智能哪一个分支领域的学习,您都应该理解它里面所有引入的概念,而且至少要有能力实现并解决其中的一些练习题或课题。
书中关于认识和建立一般性智能模型的过程如下(如果您已经对这些知识有所了解,可以略过本部分)。
AI的基本行为能力按纵向划分,大致可分为3方面:
(1)逻辑推理能力
(2)规划能力
(3)学习能力
澳大利亚国立大学在上述3个领域都有专家。从历史上看,在20世纪50年代,人工智能相关的许多具体的实际应用从机器的“逻辑推理能力”开始着手研究。然而至少在人类中,高层逻辑推理似乎是从更基本的“学习能力”和“规划能力”方面出现的,可以想象“逻辑推理能力”在一般人工智能系统中不能起到根本的、决定性的作用。
所以我将专注于“规划能力”和“学习能力”的研究。如果将这些内容相结合,在不确定性的前提下把“学习能力”、“规划能力”放在一起,就能形成“强化学习(RL)”领域的主要理论基础,这些理论在其他领域也被成为“自适应控制”或“序贯决策”理论。
Artificial Intelligence. Russell and Norvig (2003) [RN10] is the textbook to learn about Artificial Intelligence. The book gives a broad introduction, survey, and solid background of all aspects of AI. There is no real alternative. Whatever subarea of AI you specialize later, you should understand all introduced concepts, and have implemented and solved at least some of the exercises.
The textbooks below are relevant for understanding and modeling general intelligent behavior. If you already got attracted to some specific AI applications, they may not be relevant for you. One axis of categorizing AI is into (1) logical (2) planning and (3) learning aspects. CSL@ANU has experts in all 3 areas. Historically, AI research started with (1) in the 1950s, which is still relevant for many concrete practical applications. Since at least in humans, high-level logical reasoning seems to emerge from the more basic learning and planning aspects, it is conceivable that (1) will play no fundamental role in a general AI system. So I will concentrate on (2) and (3). If put together, learning+planning under uncertainty is mainly the domain of reinforcement learning (RL), also called adaptive control or sequential decision theory in other fields.
关于强化学习:
Richard S. Sutton和 Andrew G. Barto编著的《强化学习》(Reinforcement Learning,RL)被公认为是优良的RL教科书。
在学习它之前不需要具备什么背景知识。该书深入浅出的描述了一些关键思想,开放性问题,以及这个领域的大量应用示例。不要对这本“如此简单”的书感到惊讶,它主要讲解“人工智能”相关的基本知识和概念,而不是证明其在实际实现时的正确性。如果它里面详细讲解在实践中如何实现功能或者证明其理论的正确性,这会使本书非常晦涩难懂和不易学习。
至此,如果你想进一步理顺学习的方法和思路,并且出于好奇,你想更深的了解不同知识之间的关联或将现有系统功能拓展的更通用、更强大,那么,你就必须先学习一些起初看起来似乎毫不相关的、枯燥难懂的理论与概念。
Reinforcement Learning. Sutton and Barto (1998) [SB98] is the excellent default RL textbook. It requires no background knowledge, describes the key ideas, open problems, and great applications of this field. Don't be surprised about the ease of the book, it teaches understanding, not proofs. It gets really tough to make things work in practice or to prove things [BT96].
If you want to bring order into the bunch of methods and ideas you've learned so far, and want to understand more deeply their connection either for curiosity or to extend the existing systems to more general and powerful ones, you need to learn about some concepts that at first seem quite disconnected and theoretical.
关于信息理论:
智能中包含有很多信息处理过程。算法信息论(AIT)是信息论的一个分支,它有足够的能力作为智能信息处理的基础。它主要可以处理智能、相似性、创造力、类比推理和概括等方面的内容,从根本上与“归纳问题”和“奥卡姆剃刀原理”相关联。Li和itanyi编写的《算法信息论》对上述内容进行了很好的介绍。学习这本书还需要具备的基础背景知识有:Kolmogorov复杂性,最小描述长度,通用索洛莫诺夫归纳,通用莱文搜索等等。在此特别指出,这些基础背景知识,可以从经典教材《Introduction to Automata Theory, Languages, and Computation》中获取。
Information theory. Intelligence has a lot to do with information processing. Algorithmic information theory (AIT) is a branch of information theory that is powerful enough to serve as a foundation for intelligent information processing.
It can deal with key aspects of intelligence, like similarity, creativity, analogical reasoning, and generalization, which are fundamentally connected to the induction problem and Ockham's razor principle. Li and Vitanyi's (1997) AIT book [
LV97] provides an excellent introduction. Kolmogorov complexity, Minimal Description Length, universal Solomonoff induction, universal Levin search, and all that. It requires a background in theoretical computer science in general and computability theory in particular, which can be obtained from the classic textbook[HMU06].
关于通用人工智能(Artificial General Intelligence/Universal Artificial Intelligence,AGI/UAI):
我编写的这本书《Universal Artificial Intelligence Sequential Decisions based on Algorithmic Probability》从开发者的视角和完全数学化理论讲述了一种多用途的理想的通用“智能”学习代理。该书给出了这种“学习代理”的完整概念(而不是仅给出一个空空的框架)和详细定义。但请注意,这只是一个理论。它距离实际实现的差距就像是从“极大极小理论”到“一套真正的国际象棋程序”那么大,之后还有很长的路要探索。这对于从“AIXI模型”理论到真正实现“实际的通用多用途智能代理”来说是一条极为漫长和不平坦的道路。
Universal AI. Now you are in a position to read [Hut05]. The book develops a sound and complete mathematical theory of an optimal "intelligent" general-purpose learning agent. The theory is complete in the sense that it gives a complete description of this agent, not just an incomplete framework with gaps to be filled. But be warned, it is only a theory. Like it is a long way from e.g. the minimax theory of optimally playing games like chess to real chess programs, it is a long way from this theory to a practical general purpose intelligent agent [VNHS09].
下面再为大家推荐一些其他方面的数据作为扩展阅读(学习),丰富背景知识,加深各位对AI各个重要研究领域的有益补充。首先可以加入书单的是由Christopher Bishop编写的《模式识别与机器学习》(Pattern Recognition and Machine Learning),它被公认为是关于“统计机器学习”方面的优秀著作。
另外一些“贝叶斯概率”理论方面的书籍(例如:S.James Press编写的《Subjective and Objective Bayesian Statistics: Principles, Models, and Applications 2nd Edition》、 E. T. Jaynes编写的《Probability Theory: The Logic of Science 1st Edition》)也很有用,应当加入书单。
还有,博弈论领域的《Logic: An Introduction (Fundamentals of Philosophy)》(Greg Restall 编著)。《Computer Vision: A Modern Approach 1st Edition》(David A. Forsyth 编著)、《Speech and Language Processing, 2nd Edition 2nd Edition》(Daniel Jurafsky 和 James H. Martin 编著)、《Probabilistic Robotics (Intelligent Robotics and Autonomous Agents series)》(Sebastian Thrun , Wolfram Burgard 和 Dieter Fox 编著)讲述了各种理论在现实中的接口抽象定义。
最后,Nicholas Alchin编写的《Theory of Knowledge 2nd Edition》将一般性科学哲学问题及相关理论向我们娓娓道来,而John Earman编写的《Bayes or Bust? A Critical Examination of Bayesian Confirmation Theory y First edition Edition》则着重给我们介绍了“归纳问题”的相关理论知识。
Peripheral Areas. The other recommended books below can be regarded as further readings that provide more background and deepen your understanding of various important aspects in AI research. Bishop (2006) [Bis06] is the excellent default textbook in statistical machine learning, and should be put on your reading list. Some Bayesian probability book will be useful too [Pre02, Jay03]. How multiple rational agents interact [SLB08] is the domain of game theory [OR96]. Computer vision [FP02], natural language understanding [JJ08], and robotics [TBF05] interfaces abstract agents with the real world. Alchin (2006) [Alc06] gently and broadly introduces you to philosophy of science in general and Earman (1992) [Ear92] to the induction problem in particular.
Copyright © 2003-2013 www.wpsshop.cn 版权所有,并保留所有权利。