AI人工智能导论 (2).pdf

《AI人工智能导论 (2).pdf》由会员分享，可在线阅读，更多相关《AI人工智能导论 (2).pdf（23页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、Principles of Artificial Intelligence 1 History of Artificial Intelligence Principles of Artificial Intelligence : Basics : Introduction 2 The History of AI 1.3. History of Artificial Intelligence YearsDescription 19501956The Birth of AIAI的诞生 19561974The Golden Years黄金之年 19741980The First AI Winter第

2、一个AI之冬 19801987The Boom of AIAI的繁荣期 19871993The Second AI Winter第二个AI之冬 1993PresentThe Breakthrough突破 Principles of Artificial Intelligence : Basics : Introduction 3 1950, Alan Turing proposes the Turing Test as a measure of machine intelligence. 1950年，艾伦图灵提出了图灵测试，将其作为机器智能的度量。 1956, the field of Art

3、ificial Intelligence research was founded at a conference on Dartmouth College. 1956年，在美国达特茅斯学院的会议上，人工智能研究领域正式诞生。 19501956, The Birth of AI 1.3. History of Artificial Intelligence John McCarthy 约翰麦卡锡 Marvin Minsky 马文明斯基 Allen Newell 艾伦纽厄尔 Herbert Simon 赫伯特西蒙 Principles of Artificial Intelligence : B

4、asics : Introduction 4 1958, Herbert Simon and Allen Newell had a demonstration the first AI program, Logic Theorist (LT). 1958年赫伯特西蒙和艾伦纽厄尔演示了第一个AI程序，名称为逻辑理论家(LT)。 1958, John McCarthy (MIT) invented Lisp programming language. 1958年约翰麦卡锡发明了著名的Lisp编程语言。 1960s, M. Masterman and colleagues at University

5、 of Cambridge design semantic nets for machine translation. 1960年代，M马斯特曼与剑桥大学的同事们设计了语义网络，用于机器翻译。 1963, Leonard Uhr and Charles Vossler published “A Pattern Recognition Program That Generates, Evaluates, and Adjusts Its Own Operators”, which described one of the first machine learning programs. 1963年

6、伦纳德武赫和查尔斯瓦斯勒发表了关于模式识别的论文，描述了第一个机器学习程 序。 19561974, The Golden Years 1.3. History of Artificial Intelligence Principles of Artificial Intelligence : Basics : Introduction 5 1965, E. Feigenbaum initiated Dendral, a software to deduce the molecular structure of organic compounds. It was the first expert

7、 systems. 1965年，E费根鲍姆开创了Dendral，一个推断有机化合物分子结构的软件。这是首套专家系统。 19561974, The Golden Years 1.3. History of Artificial Intelligence 1974, T. Shortliffe demonstrated MYCIN program, a very practical rule-based approach to medical diagnoses. 1974年，T肖特列夫演示了MYCIN程序，一个非常实用的基于规则的医学诊断方法。 Principles of Artificial

8、Intelligence : Basics : Introduction 6 1966, the failure of machine translation. 1966年，机器翻译失败了。 1970, the abandonment of connectionism. 1970年，连接主义遭到遗弃。 19711975, DARPAs frustration with the Speech Understanding Research program at Carnegie Mellon University. 1971年至75年，美国DARPA对卡内基梅隆大学的语音理解研究项目感到沮丧。 1

9、973, the large decrease in AI research in the United Kingdom, in response to the Lighthill report “Artificial Intelligence: A General Survey”. 1973年，受莱特希尔的“人工智能：综合调查”报告的影响，英国大幅度缩减AI的研究。 19731974, DARPAs cutbacks to academic AI research in general. 1973-74，美国DARPA削减了一般性AI学术研究经费。 19741980, The First A

10、I Winter 1.3. History of Artificial Intelligence Principles of Artificial Intelligence : Basics : Introduction 7 1980, First National Conference of the American Association for Artificial Intelligence (AAAI) held at Stanford. 1980年，美国人工智能学会（AAAI）在斯坦福大学召开了第一届全国大会。 1982, Japan started Fifth Generation

11、 Computer System (FGCS) project for knowledge processing. 1982年，日本启动了第五代计算机系统（FGCS）项目，用于知识处理。 19801987, AI Boom 1.3. History of Artificial Intelligence GenerationYearsDescription 1stMid-1940sThermionic vacuum tubes真空管 2nd1956Transistors晶体管 3rd1964Integrated circuits集成电路 4th1972Microprocessors微处理器 Th

12、e Generations of Computer Principles of Artificial Intelligence : Basics : Introduction 8 In mid-1980s, the machine learning came, when the decision tree model was invented and distributed as software. The model can be viewed by a human and is easy to explain. 1980年代中期，机器学习出现了，当时发明了决策树模型 并且以软件形式推出。该

13、模型具有可视化、易说明的特点。 Also in mid-1980s, multi-layer Artificial Neural Networks (ANN) invented. With enough hidden layers, a ANN can express any function, thus overcoming the limitation of perceptron. 1980年代中期，还发明了多层人工神经元网络（ANN）。具 有足够多的隐藏层，一个ANN可以表达任意的功能，因此突 破了感知的局限性。 19801987, AI Boom 1.3. History of Art

14、ificial Intelligence Principles of Artificial Intelligence : Basics : Introduction 9 1987, the collapse of the Lisp machine market. 1987年，Lisp机的市场崩溃。 1988, the cancellation of new spending on AI by the United States governments Strategic Computing Initiative. 1988年，美国政府的战略计算促进会取消了新的AI经费。 1993, exper

15、t systems slowly reaching the bottom. 1993年，专家系统缓慢滑向低谷。 1990s, the quiet disappearance of the fifth-generation computer projects original goals. 1990年代，日本第五代计算机项目未能达到其初始目标， 悄然退场。 19871993, The Second AI Winter 1.3. History of Artificial Intelligence Principles of Artificial Intelligence : Basics : I

16、ntroduction 10 In 1997, Deep Blue became the first computer chess-playing system to beat a reigning world chess champion, Garry Kasparov. 1997年，深蓝战胜了卫冕国际象棋冠军加里卡斯帕罗夫，成为第一台计算机国际象棋系统。 1993Present, Breakthrough 1.3. History of Artificial Intelligence Deep Blue In 2005, a Stanfords Stanley, an autonomous

17、 robotic vehicle, won the DARPA Grand Challenge. 2005年，斯坦福的自主机器人车辆Stanley，赢得了DARPA无人 驾驶汽车挑战赛。 In 2006, the term “deep learning” gained traction after a publication by Geoffrey Hinton and Ruslan Salakhutdinov. 2006年，在杰弗里辛顿和鲁斯兰 萨拉赫丁诺夫在科学杂 志上发表了有关“深度学习”的论文之后，该术语成了热门。 Principles of Artificial Intelligen

18、ce : Basics : Introduction 11 Watson, the computer was specifically developed by IBM to answer questions on the quiz show Jeopardy! 沃森，是IBM专门开发的在智力竞赛Jeopardy!回答问题的计算机。 1993Present, Breakthrough 1.3. History of Artificial Intelligence In 2011, Watson competed on Jeopardy! against former winners Brad

19、Rutter and Ken Jennings. 2011年，沃森在Jeopardy!上战胜了上 届冠军布拉德路特和肯恩詹宁斯。 Watson received the first prize of $1 million. 沃森获得了1百万美元大奖。 Watson on quiz show Jeopardy! Principles of Artificial Intelligence : Basics : Introduction 12 In 2011, Google started Deep Learning project, Google Brain, as one of the Goog

20、le X projects. 2011，谷歌启动了深度学习项目，谷歌大脑，作为Google X项目之一。 1993Present, Breakthrough 1.3. History of Artificial Intelligence Google brain is a cluster of 16,000 computers dedicated to mimicking some aspects of human brain activity. 谷歌大脑是由1万6千台计算机连成的一个集群，致 力于模仿人类大脑活动的某些方面。 It had successfully recognized a

21、cat based on 10 million digital images. 通过1千万张数字图片的学习，已成功地学会识别一 只猫。 Principles of Artificial Intelligence : Basics : Introduction 13 In 2012, Siri was introduced by Apple as an integral part of iOS since iOS 5, running from iPhone 4S. 2012年，苹果公司引进了Siri，从iPhone 4S上运行的iOS5开始，已作为iOS的一个组成部分。 1993Present

22、, Breakthrough 1.3. History of Artificial Intelligence Siri is an intelligent personal assistant and knowledge navigator. Siri是一种智能个人助理和知识导航软件。 Use a natural language user interface to answer questions, make recommendations, and perform actions. 使用自然语言用户接口来回答问题、做出建议和执行动作。 Available in: English, Fren

23、ch, German, Japanese, Chinese, Korean, Italian, Spanish. 支持英语、法语、德语、日语、中文、韩文、意大利语、西班牙语。 Principles of Artificial Intelligence : Basics : Introduction 14 In 2012, Rick Rashid, Microsofts Chief Research Officer, demonstrated a real-time English-to- Chinese universal translator that keeps your voice an

24、d accent. 2012年,瑞克拉希德，微软首席研究官，演示了一款实时的英语-中文通用翻译系统，可以保持 你的声音和口音。 1993Present, Breakthrough 1.3. History of Artificial Intelligence Not only is the translation very accurate, but the software also preserves the users accent and intonation. 该软件不仅翻译非常准确，而且能够 保持讲者的口音和语调。 Principles of Artificial Intellig

25、ence : Basics : Introduction 15 Apr. 2014, Microsoft demonstrated “Cortana”, an intelligent personal assistant on Windows Phone. 2014年4月，微软演示了“Cortana”，一款运行在Windows Phone上的智能个人助理。 Jun. 2014, Microsoft China released chatbot “XiaoIce (小冰)” which allowed WeChat users to have conversations with it. 201

26、4年6月，微软中国推出了聊天机器人小冰，可让微信用户与她交谈。 1993Present, Breakthrough 1.3. History of Artificial Intelligence Cortana 小冰 度秘 2015年9月8日，百度在2015百度世界大 会上推出了一款机器人助理度秘，可 以为用户提供秘书化搜索服务。 Principles of Artificial Intelligence : Basics : Introduction 16 Jun. 2014, chatbot Eugene Goostman, at a contest marking the 60th an

27、niversary of Turings death, 33% of the events judges thought that Goostman was human, so that the events organizer considered it to have passed Turings test. 2014年6月，聊天机器人尤金古斯特曼，在纪念图灵逝世60周年的一个比赛上，被该活动33% 的评委认为古斯特曼是人类，因此组织者认为它已经通过了图灵测试。 1993Present, Breakthrough 1.3. History of Artificial Intelligenc

28、e Eugene Goostman is developed in Saint Petersburg in 2001 by a group of three programmers. 尤金古斯特曼是由三个程序员小组于 2001年在圣彼得堡开发的。 Principles of Artificial Intelligence : Basics : Introduction 17 Aug. 2014, IBM announced “TrueNorth” chip to work like human brain. 2014年8月，IBM发表了类人脑工作的TrueNorth芯片。 1993Presen

29、t, Breakthrough 1.3. History of Artificial Intelligence TrueNorth is a neuromorphic CMOS chip, consists of 4096 hardware cores, each one simulating 256 programmable silicon neurons for a total of just over a million neurons. TrueNorth是一款神经形态的CMOS芯片， 由4096个硬件核组成，每个仿真256个可 编程的硅神经元，总计刚好超过百万个 神经元。 Princ

30、iples of Artificial Intelligence : Basics : Introduction 18 Feb. 2015, Google DeepMind published Deep Q-Network, the human-level control through deep reinforcement learning. 1993Present, Breakthrough 1.3. History of Artificial Intelligence Tested on classic Atari video games (late-1970s and early-19

31、80s) 在经典（1970年代后期至1980年代初）的Atari视频游戏机上进行的测试 At human-level or above (29/49 59.18%) Below human-level (20/49 40.82%) 2015年2月，谷歌DeepMind公司在Nature杂志上发表了Deep Q-Network，通过深度强化 学习达到人类水平的操控。 Principles of Artificial Intelligence : Basics : Introduction 19 Dec. 2015, Google DeepMinds program AlphaGo beat Fa

32、n Hui, the European Go champion, five times out of five in tournament conditions. 2015年12月，谷歌DeepMind公司的程序AlphaGo打败了欧洲围棋冠军樊麾，成绩5战5胜。 1993Present, Breakthrough 1.3. History of Artificial Intelligence Jan. 27 2016, the announcement of the news was delayed until this day, to coincide with the publicati

33、on of a paper in the journal Nature describing the algorithms used. 这个消息直到2016年1月27日才宣布，目的是与描述所用算法 的论文在自然杂志发表的时间同步。 Deep-learning software defeats human professional for first time. 深度学习软件第一次击败了人类职业棋手。 Principles of Artificial Intelligence : Basics : Introduction 20 Mar. 9-15 2016, AlphaGo played So

34、uth Korean professional Go player Lee Sedol, ranked 9-dan, in Seoul, South Korea. AlphaGo won all but the fourth game. 2016年3月8日至15日，AlphaGo在韩国首尔对垒韩国九 段职业棋手李世乭。AlphaGo以5战4胜赢得了比赛。 1993Present, Breakthrough 1.3. History of Artificial Intelligence Source: http:/www.goratings.org/history/ Principles of

35、Artificial Intelligence : Basics : Introduction 21 2016年4月9日，我是歌手第四季总决赛落下帷幕，李玟夺得总冠 军。据报道，在决赛结果宣布之前阿里云小Ai就预测到了李玟夺冠。 1993Present, Breakthrough 1.3. History of Artificial Intelligence 小Ai是阿里云研发的人 工智能程序，主要基于 神经网络、社会计算、 情绪感知等原理工作， 善于洞察本质和实时预 测，并能理解人类情感 ，可以通过强大的计算 和机器学习能力不断自 我进化。 Principles of Artificial

36、Intelligence : Basics : Introduction 22 Jan. 19 2016, the Information Technology and Innovation Foundation (ITIF) in Washington DC announced its annual Luddite Award. 2016年1月19日，位于华盛顿特区的信息技术与创新基金会(ITIF)，公布了年度卢德奖。 ITIF gave the Luddite Award to “a loose coalition of scientists and luminaries who stir

37、red fear and hysteria in 2015 by raising alarms that artificial intelligence (AI) could spell doom for humanity”. ITIF将该卢德奖颁发给“一个科学家和名人组成的松散联盟，他们在2015年警告人工智能(AI) 将会导致人类的末日，激起恐惧和歇斯底里”。 1993Present, Breakthrough 1.3. History of Artificial Intelligence Source: https:/itif.org/publications/2016/01/19/ The winners if that is the correct word included Elon Musk, and Stephen Hawking. 获奖者如果这个词没用错的话包括埃隆马斯克 、和史蒂芬霍金。