《人工智能英文版课件：01 intoduction》由会员分享，可在线阅读，更多相关《人工智能英文版课件：01 intoduction（119页珍藏版）》请在金锄头文库上搜索。
1、Artificial IntelligenceCourse OutlinenGrading SchemenAttendance (10%)nAssignment (20%)n2 assignmentsnMid-term Test (25%)nAround 9th weeknFinal Exam (45%)Scope of 2013 Artificial IntelligencenIntroduction of AI (Chapter 1)nIntelligent Agents (2)nSolving Problem by Searching (3)nBeyond Classical Searc
2、h (4)nAdversarial Search (5)nConstraint Satisfaction Problems (6)nLearning from Examples (18)nPlanning (10)nCurrent Research Topics (Pending)nDealing with Big Data, e.g. Image RetrievalnDeep Learning for Image ClassificationChapter 1 Introduction of Artificial intelligenceOutlinenWhy we study AInWha
3、t is AI?nThe Foundations of AInThe History of artifical intelligencenThe State of ArtnSummaryWhy we study AInHuman a.k.a. Homo Sapiens: Latin of Wise HumannMental Capacities learn, predict, create, etcnSense of SelfnAI address one of the ultimate puzzlesnHow is it possible for a slow, tiny brain to
4、perceive, understand, predict and manipulate a world far larger and more complicated than itself?nArtificial Intelligence or AInTry to understand intelligent entitiesnTry to build intelligent entities (unlike philosophy & psychology)nA relative new research area started in 1956nWith a large influenc
5、e to many other research areasnPattern Recognition, Machine Learning, Robotics, Controls, Man-Machine Interface, etcDancingPlaying footballR2-D2 and 3CPOand?!Why we study AInIntelligent Game PlayingAI in GamesWhy we study AInPattern recognitionImage ClassificationStock Market Candlestick PatternsHan
6、dwritten charactersand Why we study AInIntelligent Traffic ControlWhy we study AInSearch enginesWhy we study AInGoogle TranslationnTranslate whole paragraphnPowerfulreally?What is AI?nThere is no solid definition of Artificial, the definition differs for ndifferent peoplendifferent contextsndifferen
7、t historical periods. nSome definitions of AI organized into four categories 1.Systems that think like humans. 2.Systems that think rationally. 3.Systems that act like humans. 4.Systems that act rationally. What is AI?The exciting new effort to make computers thinks machine with minds, in the full a
8、nd literal sense” (Haugeland 1985)“The art of creating machines that perform functions that require intelligence when performed by people” (Kurzweil, 1990)“The study of mental faculties through the use of computational models” (Charniak et al. 1985)A field of study that seeks to explain and emulate
9、intelligent behavior in terms of computational processes” (Schalkol, 1990)Systems that think like humansSystems that think rationallySystems that act like humansSystems that act rationallyWhat is AI?nThe four definitions above vary along two dimensions :nHuman-centerednRationalitynHuman-centered app
10、roachnempirical science, involving hypothesis and experimental confirmation.nRationalist approachncombination of mathematics and engineering.What is AI?nActing humanly: Turing Test Alan Turings 1950 article Computing Machinery and Intelligence proposed Turing Test,which provide a satisfactory operat
11、ional definition of intelligence. Computing Machinery and IntelligenceWhat is AI? Turing Test Turing TestWhat is AI? Computing Machinery and Intelligencen“Can machines think?” “Can machines behave intelligently?nThe Turing test (The Imitation Game): nA human interrogator poses written questions to i
12、tnThe computer return written responsesnThe computer passed the test if the human interrogator can not tell whether the responds are from a person or notWhat is AI?nComputer needs to possess: nNatural language processingnKnowledge representationnAutomated reasoningnMachine learning. nTuring test avo
13、ided direct physical interaction between the interrogator and the computer.nDisadvantages of the Turing test nnot reproduciblennot constructivennot amenable to mathematical analysis.What is AI?nAI researchers have devoted little effort to passing the Turing testnBut devoted much effort to study the
14、underlying principle of intelligence.nFor example, to build a flying machine, we do not need to imitate birds but go to learn aerodynamics.n“artifical flight” succeeded. Artificial versus Natural FlightWhat is AI? What does the C-3PO need to communicate with human?What is AI?What does the C-3PO need
15、 to communicate with human? nNatural language processing to enable it to comminicate successfully in English.nKnowledge representation to store what it knows or hears.nAutomated reasoning to use the stored information to answer questions and to draw new conclusions.What is AI?nMachine learning to ad
16、apt to new circumstances and to detect and extrapolate patterns.nComputer Vision to recognize the interrogators actions and various objects presented by the interrogator.nRobotics to manipulate objects and move aboutWhat is AI?nThinking Humanly: Cognitive modelingnHow humans think? To determine how
17、humans think, we need to get inside the actual working mechanism of human mind.There are two way to do this:nThrough introspection nThrough psychological experiment nA program behave humanly If a programs input/output and timing behaviors match corresponding human behaviorsWhat is AI?The Neural comp
18、uter is able to imitate human brainWhat is AI?n1960 “Cognitive Revolution”ninformation-processing psychology replaced prevailing orthodoxy behaviorism.nRequires scientific theories of internal activities of the brainnWhat level of abstraction? “Knowledge” or “Circuits”?nHow to validate models?nPredi
19、cting and testing behavior of human subjects (top-down)nDirect identification from neurological data (bottom-up)nBuilding computer/machine simulated models and reproduce results (simulation)What is AI?nBoth approaches(roughly,Cognitive Science and Cognitive Neuroscience) are now distinct from AI.nBo
20、th share with AI the following characteristic:nthe available theories do not explain (or engender)anything resembling human-level general intelligence nHence, all three fields share one principle direction!What is AI?nThinking Rationally: Laws of Thought nAristotle: What are correct arguments/though
21、ts processes? ncodify “right thinking”-a reasoning processes nAristotles famous syllogism provide patterns for argument structures. “Socrates is a man, all men are mortal; therefore Socrates is mortal” nStudying the law of thought initiated the field called logic What is AI?A famous Greek philosophe
22、r: AristotleWhat is AI?nSeveral Greek schools developed various forms of logic:nnotation and rules of derivation for thoughts.nmay or may not have proceeded to the idea of mechanizationnAll kinds of things in the world and their relations can be stated by a precise notation.nBy 1965, programs existe
23、d that could, given enough time and memory, take a description of a problem in logical notation and find the solution to the problem if one existsnLogicist tradition within AI hopes to build on such programs to create intelligent systems What is AI?nproblemsnNot all facts are cerntainnPrinciple some
24、times disagrees with practiceWhat is AI?nActing Rationally: The Rational Agent 1.Acting rationally nRational behavior: doing the right thing nThe right thing: is expected to maximize goal achievement for given and available information What is AI?nDoes not necessarily involve thinkingnblinking refle
25、x-but thinking should be in the service of rational action. nAristotle (Nicomachean Ethics)nEvery art and every inquiry, and similarly every action and pursuit, is thought to aim at some goodWhat is AI? 2. Rational agentsnAn agent Commonly an agent is something that acts. Abstractly, an agent is a f
26、unction from percept histories to actions: .f:p*-A But an computer agent is an entity that have other attributes, such as operating under autonomous control, perceiving their environment, etcWhat is AI?nRational agentsnIt is one that acts so as to achieve either the best outcome or the best expected
27、 outcome when there is uncertaintynFor any given class of environments and tasks, we seek the agent (or class of agents) with the best performancenComputational limitations make perfect rationality unachievablenWhat we can do is to design the best program for given machine resources.The Foundations
28、of AInAI is the interdisciplinary study of science includingnPsychologynPhilosophynNeurosciencenMathematicsnLinguisticsnThey contributed ideas, viewpoints and techniques to AI. nThey are the foundations of AI.nWe organize these foudations around questions related to AI and contributed methods,result
29、s to AI. nOperations researchnControl theorynCyberneticsnEconomicsnComputer EngineeringThe Foundations of AInPhilosophy(428B.C.-present) Questions Can formal rules be used to draw valid conclusions? How does the mental mind arise from a physical brain? Where does knowledge come from? How does knowle
30、dge lead to action? Contributed methods and results Logic, methods of reasoning, Mind as physical system Foundations of learning, language, rationalityThe Foundations of AInMathematics(c.800- present) Questions what are the formal rules to draw valid conclusion? what can be computed? How do we reaso
31、n with uncertain information? Contributed methods and results Formal representation and proof Algorithms Computation, (un)decidability, (in)tractability, probabilityThe Foundations of AInEconomics(1776- present) Questions How should we make decisions so as to maximize payoff? How should we do this w
32、hen others may not go along? How should we do this when the payoff may be far in the future? Contributed methods and results utility decision theoryThe Foundations of AInNeuroscience(1861- present) Questions How do brain process information? Contributed methods and results physical substrate for men
33、tal activityThe Foundations of AInPsychology(1879- present) Questions How do humans and animals think and act? Contributed methods and results phenomena of perception motor control experimental techniquesThe Foundations of AInControl theory and Cyberneties(1948- present) Questions How can we build a
34、n efficient computer? Contributed methods and results design systems that maximize an objective function over timeThe Foundations of AInLinguistics(1957- present) Questions How can artifacts operate under their own control? Contributed methods and results knowledge representation grammarThe Foundati
35、ons of AInComputer Engineering(1940-present) Questions How does language relate to thought? Contributed methods and results building fast computersThe History of artifical intelligenceTimeline of major AI events The History of artifical intelligencenEvidence of AI folklore can be traced back to anci
36、ent EgyptnWith the development of the electronic computer in 1941, the technology finally became available to create machine intelligence. nThe term artificial intelligence was first coined in 1956, at the Dartmouth conferencensince then Artificial Intelligence has expanded because of the theories a
37、nd principles developed by its dedicated researchers. nThrough its short modern history, advancement in the fields of AI have been slower than first estimated. The History of artifical intelligencenProgress continues to be made. nFrom its birth 6 decades ago, there have been a variety of AI programs
38、nThey have impacted other technological advancements. The History of artifical intelligenceHistory list nPrecursorsnThe gestation of AI(1943-1955)nThe birth of AI(1956)nEarly enthusiam,great expectation(1952-1969)nA dose of reality(1966-1973)nKnowledge-based systems:The key to power?(1969-1979)nAI b
39、ecomes an industry(1980-present)nThe return of neural network(1986-present)nAI becomes a science(1987-present)nThe emergence of intelligent agents(1995-present)nAvailability of Very Large Data Sets (2001-present)The History of artifical intelligencen Precursors The 1949 innovation of the stored prog
40、ram computer made the job of entering a program easier Advancements in computer theory lead to computer science, and eventually Artificial intelligence. The History of artifical intelligenceAl-Jazaris programmable automata AutomatonsFormal reasoningComputer scienceThe History of artifical intelligen
41、cenThe gestation of AI (1943-1955)n1943 McCulloch&Pitts The first AI work: artificial neural net, proved equivalent to Turing machine n1949 Donald Hebb Hebbian learning, demonstrated a simple updating rule for modifing the connection strengths between neurons-remains an influential model to this day
42、sThe History of artifical intelligencen1950 Alan TuringnHis article “Computing Machinery andnIntelligence” is the first complete vision of AI.nIn this article,he introduce the Turing test, machine learning, genetic algorithms and reinforcement learning.n1951 Minsky&Edmonds SNARC,the first neural net
43、work computerThe History of artifical intelligencenThe birth of Artificial intelligence(1956) Dartmouth Summer Research Conference on Artificial Intelligence in 1956 symbolizes the birth of AI. The birthplace of AI: Dartmouth CollegeThe History of artifical intelligence nJohn McCarthy coined the ter
44、m “AI” Why it is AI ? Perhaps “Computational rationality” would be better.n Newell & Simon presented LOGIC THEORIST (LT) program It is a reasoning program, capable of thinking non- numerically. It could prove most of theorems like Russell and Whiteheads Principia Mathematica.The History of artifical
45、 intelligence Why AI become a separate field? 1. AI from the start embraced the idea of duplicating human faculties, self-improvement and language use. None of the other fields were addressing these issues. 2. AI is the only one of these fields that is clearly a branch of computer science and AI is
46、the only field to attempt to build machines that will function autonomously in complex & changing environments. The History of artifical intelligencenJohn McCarthy nIn 1956, John McCarthy regarded as the father of AI, organized a conference to draw the talent and expertise of others interested in ma
47、chine intelligence for a month of brainstorming. nHe invited them to Vermont for The Dartmouth summer research project on artificial intelligence. nFrom that point on, because of McCarthy, the field would be known as AI.nAlthough not a huge success, the Dartmouth conference did bring together the fo
48、unders in AI, and served to lay the groundwork for the future of AI research. The father of AI: John McCarthy The History of artifical intelligencenEarly enthusiasm, great expectations(1952-1969) n1957 Herb SimonnIt is not my aim to surprise or shock you but the simplest way I can summarize is to sa
49、y that there are now in the world machines that think, that learn and that create. nMoreover their ability to do these things is going to increase rapidly untilin the visible futurethe range of problems that can handle will be coextensive with the range to which human mind has been applied.The Histo
50、ry of artifical intelligencen1958 John McCarthy defined the high-level language Lisp, which is the second-oldest major high-level language in cerrent use.n1965 : J.A. Robinson invents the resolution principle, basis for automated theorem provingnA machine can never do XnPart of the list of X include
51、: fall in love, enjoy food, diversified behavior like humanThe History of artifical intelligencenIntelligent reasoning in Microworlds (such as Blocks world)nMarvin Minsky pointed out that: successful sciences were often best understood using simplified models like frictionless planes or perfectly ri
52、gid bodies. nMuch of the research focused on the so-called blocks world, which consists of colored blocks of various shapes and sizes arrayed on a flat surface .The History of artifical intelligence The Blocks worldThe History of artifical intelligencenA dose of reality(1966-1973)n1965 : Weizenbaums
53、 ELIZAnDifficulties in automated translation, n“the spirit is willing but the flesh is weak”n - “the vodka is good but the meat is rotten”nLimitations of Perceptrons discovered (XOR problem)nMachine evolution (now Genetic Algorithms)nSystems for microworlds dont scale up for real applications XOR pr
54、oblemnSeparate and nPerceptron separates samples in two class by a straight line (y=AX)nNo single straight line can do itnPerceptron fails this simple XOR problemThe History of artifical intelligencen1973 Lighthill report -cut AI fundingnEarly AI systems turned out to fail miserably when tried out o
55、n wider selection of problem and on more difficult problems:1.most early programs contained little or no knowledge of their subject matter, they succeeded by means of simple syntactic manipulations. 2.the intractability of many of the problem that AI was attempting to solve3.some fundamental limitat
56、ion on the basic structures being used to generate intelligence behaviorThe History of artifical intelligencenKnowledge-based systems: The key to power? (1969-1979)nweak methodsnA geneal-purpose search mechanism trying to string together elementary reasoning steps to find complete solutions.nWhy nee
57、d Knowledge?nWeak methods do not scale up to large or difficult problem instances.nThe alternative to weak methods is to use more powerful domain-specific knowledgenallows larger reasoning stepsnmore easily handle typically occuring cases in narrow areas of expertise.The History of artifical intelli
58、gencenDendral The first successful knowledge-intensive system. Inferring molecular structure from the information provided by a mass spectrometernThe Heuristic Programming Project(Hpp) to investigate the extent to which the new methodology of expert systems could be applied to other areas of human e
59、xpertise.nMYCIN to diagnose blood infections.With about 450 rules.MYCIN was able to perform as well as some experts, and considerably better than junior doctorsThe History of artifical intelligencenWinograds SHRDLU system Knowledge used in the area of understanding natural language. It was able to o
60、vercome ambiguity and understand pronoun references, but mainly because it was designed specifically for one areathe blocks world. nMany programs would understanding natural language. But all on specific area,such as representing stereotypical situations,describing human memory organizationThe Histo
61、ry of artifical intelligenceThe widespread growth of application to realworld caused a concurrent increase in the demands for workable knowledge representation schemes.A large number of different representation and reasoning language were developed: Prolog PLANNER. The History of artifical intellige
62、ncenAI becomes an industry(1980 - now)nThe first successful commeicial expert system R1 began operation at the Digital Equipment CorporationnIt helps configure order for new computer system.nBy1986, it saved the company about $40million a year. nNearly every major U.S. corporation had its own AI gro
63、up and was either using or investigating expert systems.The History of artifical intelligencenFifth generation computer system projectnLogic as the basis for computing.nIn response US formed MCC as a research consortium designed to assure national competitive.nBut both MCC and the Fifth Generation p
64、roject never meet their ambitios goals. nSo the Alvery reinstated the funding that was cut by the Lighthill report. nFrom 1980 to 1988 AI industry boomed.nExpert systems emphasised : nknowledge representation rules,frames, semantic nets.The History of artifical intelligencenProblems: nthe knowledge
65、becomes a bottleneck;nmarrying expert system & traditional software;nbreaking into the mainstream.nLater 1980s many companies failed to deliver on extravagant promise, therefore the period of “AI winter” came. The History of artifical intelligencenThe return of neural networks(1986 - now) A Neural n
66、etworksThe History of artifical intelligencenThe earliest work in neural computing goes back to the 1940s when McCulloch and Pitts introduced the first neural network computing model.nSince then, many researches were made.nBut in the late 1970s it had been largely abandoned because of its limitation
67、snIn the early 1980s Neural networks returnedThe History of artifical intelligencenEarly stagesn1943 McCulloch-Pitts: nneuron as computing elementn1948 Wiener: ncyberneticsn1949 Hebb: nlearning rulen1958 Rosenblatt: nperceptronn1960 Widrow-Hoff: nleast mean square algorithmThe History of artifical i
68、ntelligencenRecessionnIn the 1950s, Rosenblatts work resulted in a two-layer network, the perceptrons, which was capable of learning certain classifications by adjusting connection weightsnAlthough the perceptron was successful in classifying certain patterns, it had a number of limitations. nIn 196
69、9 Marvin Minsky and Seymour Papert showed in their thesis, the perceptron was not able to solve the simple classic XOR (exclusive or) problem. nSuch limitations led to the decline of the field of neural networks.XOR problemnSeparate and nPerceptron separates samples in two class by a straight line (
70、y=AX)nNo single straight line can do itnPerceptron failsThe History of artifical intelligencenRevivaln1982 Hopfield: recurrent network modeln1982 Kohonen: self-organizing mapsn1986 Rumelhart : back-propagation (BP)nMultilayer Perceptron Neural Networks (MLPNN)nBack-propagation was applied to many le
71、arning problems in the computer science and psychology, and the widespread dissemination of results in the collection Parallel Distributed Processing, caused great excitement. XOR problemnSeparate and nMLPNN separates samples in two class by multiple straight lines (y=AX)nMLPNN solves XOR problem !n
72、It spends 17 years from Perceptron to MLPNN !The History of artifical intelligencenAI become a science(1987-present)nRecent years a revolution in both the content and the methodology of work in AI have happened.nNow it is more common to build on existing theories than to propose brand new ones, to b
73、ase claims on rigorous theorems or hard experimental evidence rather than on intuition, and to show relevance to real-world application rather than toy example. nAI did not resist against the limitation of existing fields, but embrace those fields. nIn term of methodology, AI has finally come firmly
74、 under the scientific method.The History of artifical intelligencenExamples:nApproaches based on Hidden Markov Models (HMMs) dominate the field of speech recognitionsnSpeech technology and the related field of handwritten character recognition are already making the transition to widespread industri
75、al and consumer application. nMany researches on Neural network were done. nUsing improved methodology and theoretical frameworksnNeural networks could be applied easily to any application if the application could be formulated as either a classification problem or a function approximation problemnN
76、ot good in dealing with uncertaintyThe History of artifical intelligencenResurgence of probabilistic and decision-theoretic methods in AI.nBayesian network nAllow efficient representation of and rigorous reasoning with uncertain knowledge.nThis approach largely overcome many problems of the probabil
77、istic reasoning systems of the 1960s and 1970s. nIt now dominates AI research on uncertain reasoning and expert systems. nSimilar gentle revolution have occurred in robotics, computer vision and knowledge representation.The History of artifical intelligencenThe emergence of intelligent agents (1995-
78、present)nIn 1990s, researchers emphasized on understanding the interaction between agents and environments.nSo an agent is a system that is situated in an environment and is capable of perceiving its environment and acting in it to satisfy its design objectives.The History of artifical intelligenceE
79、nvironmentAgent?ActuatorSensorThe History of artifical intelligencenHuman “agent”:nenvironment: physical world;nsensors: eyes, ears, . . .neffectors: hands, legs, . . .nSoftware agent:nenvironment(e.g.)UNIX operating system;nsensor:ls,ps,.neffectors:rm,chmod,The History of artifical intelligencenInt
80、ernet agent:nenvironment:the Internet;nsensor:http requests;neffectors:http commands. nAgents are widely used.nOne of the most important environments for intelligenct agents is the Internet.The History of artifical intelligencenThe consequence for the emergence of intelligent agents are the followin
81、g:nThe realization that the previosly iosolated subfields of AI might need to be reorganized somewhat when their results are to be tired together . nAI has been draw into much closer contact with other fields, such as control theory and economics, that also deal with agents.The History of artifical
82、intelligencenVery Large Datasets (2001-present)nAI focuses on Algorithm to improve intelligent in old daysnHowever, we should pay more attention to data in applicationsnLarge variety of available AI algorithmsnHow to collect and select data is importantnIf data collected is irrelevant or noisy, AI a
83、lgorithm can not learn and performThe History of artifical intelligencenCan algorithm process large datasets?nBillions of images for image retrieval on WebnTrillions of words of EnglishnBillons of base pairs of genomic sequencesnGiven the word “Plant”, do you think it is an factory or a flora?nIf a
84、few examples is available, we can label the training samples by humannIf there are 10,000 sentences with the word plant, do you want to label them by yourself?The History of artifical intelligencenGiven the word “Plant”, do you think it is an factory or a flora?nWe could use AI algorithm to learn fr
85、om a large set of examples with the help of dictionarynwork, industrial plantflora, plant lifenYatowskys work achieved 96% accuacynBanko and Brill proposed a technique that perform even better when the number of text increases to a billionnWhen the number of samples is large enough, a mediocre algor
86、ithm could perform wellThe History of artifical intelligencenYou want to Photoshop all photos to replace the face of your ex-boyfriend with the background of the imageni.e. you want to make him disappear from all of your photo, just like never existed.nHays and Efros proposed an algorithm to search
87、through a collection of photos to find something that will matchnThen, you could erase “him”nThe algorithm perform poorly when there is only 10000 imagesnThe algorithm works well when there are 2 million photosThe History of artifical intelligencenWorks like this suggests that the “knowledge bottlen
88、eck” in AInHow to express all the knowledge that a system need?nIt is better solved by learning methods instead of hand-coded knowledge, if enough data is providedDr. Ngs current worksnImage Retrieval from millions of imagesn1 Ph.D., 3 master, 2 undergraduaten3 journals and some conf. papers in 2013
89、nLearning from images with missing partsn1 Ph.D.n2 journals in 2013nLearning for imbalance and changing datan1 Ph.D., 3 masters, 3 undergraduaten2 journals and some conf. papers in 2013The History of artifical intelligencenAfter the “AI Winter”nThere may be a new AI Spring upcomingnHavenstein 2005nT
90、oday, many thousands of AI applications are embedded in the infrastructure of every industrynKurzweil 2005The State of ArtIt proved that AI is useful in many subfields.Here are several application samples:nAutonomous planning and scheduling NASA Remote Agent in Deep Space I probe explores solar syst
91、em. Remote Agent,developed at NASA Research Center and JPL, was the first artificial intelligence control system to control a spacecraft without human supervision.The State of ArtLaunch of Deep Space 1Deep Space 1s Remote Agent The State of ArtnDiagnosis In 1991 Medical diagnosis programs based on p
92、robabilistic analysis have been able to perform at the level of an expert physician in several area of medicine. Now AI is widely used. And various intelligent medical diagnosis equipments were produced. The State of ArtA Cardiovascular disease diagnosis apparatusThe State of Art Now intelligent dia
93、gnosis is also used in many other industry,such as electric-power, telecom, spaceflight. Many diagnosis systems were also produced in those industry.The State of ArtA electric diagnosis systemThe State of ArtnGame playing Deep Blue defeated the reigning world chess champion Garry Kasparov in 1997.De
94、ep Blue was a computer developed by IBM to beat Garry Kasparov, considered by some to be the greatest chess player ever. Deep Blue won this game, but Kasparov rebounded over the following 5 games to 3 wins and 2 draws, soundly beating the machine in the 1996 match. In the 1997 rematch, Deep Blue man
95、aged to win two games to Kasparovs one, taking the match 3.5 to 2.5. The State of ArtGarry Kasparov vs Deep Blue The State of ArtThe State of Artn Autonomous control The ALVINN computer vision system was trained to steer a car to keep it following a lane.It was placed in CMUs NAVLAB computer-control
96、led minivan and used to navigate across the United States. (driving autonomously 98% of the time from Pittsburgh to San Diego) The State of ArtUnmanned auto The State of Art DARPA grand challenge: Autonomous vehicle navigates across desert.The State of ArtDARPA grand challengeThe State of ArtnRobtic
97、s Many surgeons now use robot assitants in micosurgery. HipNav is a system that use computer vision techniques to creat a three-demensional model of a patients internal anatomy and then used robotic control to guide the insertion of a hip replacement prosthesis.The State of Art iRobot Roomba: automa
98、ted vacuum cleanerThe State of ArtnLogistic Planning Dynamic Analysis and Replanning Tool(DART) It can do automared logistic planning and scheduling for transportation.This involved up to 50,000 vehicles, cargo, and people at a time.The State of ArtnLanguage understanding and problem solving PROVERB
99、 Littman A computer program that solves crossword puzzles better than most humans,using constraints on possible word fillters,a large database of past puzzles,and a variety of information sources including dictionaries and online database such as a list of movies and the actors that appear in them.
100、The State of ArtThere are just a few examples of artificial intelligenceSystems above.We could learn more from the Website,Intelligent Agent,Machine Learining, Proving Theorem.AI robot around usWe are surrounded by AI applications and machines and you can not live without them totallyBefore the end
101、of todayA final question before the end of the first class of AIWe are surrounded by AI applications and machines and you can not live without them totallyAre they friendly to you? what if not?SummarynExplained why we need to study AI from the point that AIs functions.nDifferent people think AI diff
102、erently. Generally, there are four kind of AI definitions.nDifferent disciplines made respective contribution to AI. nThe history of AI has had cycles of success, misplaced optimism and resulting cutbacks in enthusiasm and funding. There have also been cycles of introduing new creative approaches and systematically refining the best ones.nRecently great achievement have been made in AI. Thank you
