A tifi i l I t lli A It d ti A tifi i l I t lli A It d ti Artificial Intelligence: An Introduction Artificial Intelligence: An Introduction Revised September 2008 Revised September 2008 Byoung-Tak Zhang School of Computer Science and Engineering School of Computer Science and Engineering Graduate Programs in Cognitive Science, Brain Science, and Bioinformatics Seoul National University Seoul National University E-mail: [email protected]http://bi snu ac kr/ http://bi.snu.ac.kr/
74
Embed
A tifi i l I t lli A I t d tiArtificial Intelligence ... - SNUA tifi i l I t lli A I t d tiArtificial Intelligence: An Introduction Revised September 2008 Byoung-Tak Zhang School of
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A tifi i l I t lli A I t d tiA tifi i l I t lli A I t d tiArtificial Intelligence: An IntroductionArtificial Intelligence: An Introduction
Revised September 2008Revised September 2008
Byoung-Tak Zhang
School of Computer Science and EngineeringSchool of Computer Science and EngineeringGraduate Programs in Cognitive Science, Brain Science,
and BioinformaticsSeoul National UniversitySeoul National University
Mars Rover Sojourner:Mars Rover Sojourner: M P hfi d Mi iM P hfi d Mi iMars Rover Sojourner: Mars Rover Sojourner: Mars Pathfinder MissionMars Pathfinder Mission
What Is Artificial Intelligence (AI)?What Is Artificial Intelligence (AI)?What Is Artificial Intelligence (AI)?What Is Artificial Intelligence (AI)?
Branch of computer science that is concerned with the automation of intelligent behavior.Design and study of computer programs that behave intelligentlyes g a d study o co pute p og a s t at be ave te ge t yStudy of how to make computers do things at which, at the moment, people are better.Designing computer programs to make computers smarterDesigning computer programs to make computers smarter.Develop programs that respond flexibly in situations that were not specifically
) h l i beg.) - house-cleaning robots- perceive its surroundings- navigate on the floor- respond to events- decide what to do next- space exploration (Fig. 1.1)p p ( g )
What is Artificial Intelligence?What is Artificial Intelligence?What is Artificial Intelligence?What is Artificial Intelligence?AI i ll i f h d bl hi h b l d bAI is a collection of hard problems which can be solved by humans and other living things, but for which we don’t have good algorithms for solving.g g g4e. g., understanding spoken natural language, medical diagnosis,
circuit design, learning, self-adaptation, reasoning, chess playing, proving math theories etcproving math theories, etc.
Definition from R & N book: a program that4Acts like human (Turing test)4Thinks like human (human-like patterns of thinking steps)4Acts or thinks rationally (logically, correctly)
Some problems used to be thought of as AI but are nowSome problems used to be thought of as AI but are now considered not4e. g., compiling Fortran in 1955, symbolic mathematics in 1965,
History of AIHistory of AIHistory of AIHistory of AIEarly enthusiasm (1950’s & 1960’s)4 Turing test (1950)4 1956 Dartmouth conference4 Emphasize on intelligent general problem solvingp g g p g
Emphasis on knowledge (1970’s)4 Domain specific knowledge4 DENDRAL, MYCIN
AI became an industry (late 1970’s & 1980’s)AI became an industry (late 1970 s & 1980 s)4 Knowledge-based systems or expert systems4 Wide applications in various domains
Searching for alternative paradigms (late 1980’s - early 1990’s)44 AI’s Winter: limitations of symbolic/logical approaches4 New paradigms: neural networks, fuzzy logic, genetic algorithms, artificial life
Resurge of AI (mid 1990’s – present)4 Internet Information retrieval data mining bioinformaticsInternet, Information retrieval, data mining, bioinformatics4 Intelligent agents, autonomous robots
Recent trends:4 Probabilistic computation4 Biological basis of intelligence
Applications of Intelligent Agents (1)Applications of Intelligent Agents (1)Applications of Intelligent Agents (1)Applications of Intelligent Agents (1)
E il AE-mail Agents4Beyond Mail, Lotus Notes, Maxims
h d liScheduling Agents4ContactFinder
Desktop Agents4Office 2000 Help, Open Sesame
Web-Browsing Assistants4WebWatcher, Letizia
Information Filtering Agents4Amalthaea, Jester, InfoFinders, Remembrance agent,
Applications of Intelligent Agents (2)Applications of Intelligent Agents (2)Applications of Intelligent Agents (2)Applications of Intelligent Agents (2)
N i ANews-service Agents4NewsHound, GroupLens, FireFly, Fab, ReferralWeb,
Evolutionary ComputationEvolutionary Computation: : y py pNature as ComputerNature as Computer
“Owing to this struggle for life, any variation, however slight and from whatever cause proceeding, if it be in any degree profitable to an individual of any species, in its infinitely complex relations to other organic beings and to external nature, will tend to the preservation of that individual, and will generally be inherited by its offspring.” p g
CC l i S S ftb t (1)l i S S ftb t (1)Application Example 14Application Example 14
CoCo--evolving Soccer Softbots (1)evolving Soccer Softbots (1)CoCo evolvingevolvingCoCo--evolvingevolvingSoccer Softbots Soccer Softbots With Genetic With Genetic P iP i
At R b C th t "l " th " l" b t
ProgrammingProgrammingAt RoboCup there are two "leagues": the "real" robot league and the "virtual" softbot leagueHow do you do this with GP?How do you do this with GP?4GP breeding strategies: homogeneous and heterogeneous4Decision of the basic set of function with which to evolve players
Computing Power and Memory Capacity of Computing Power and Memory Capacity of p g y p yp g y p yComputers and Organisms [Moravec, 1988]Computers and Organisms [Moravec, 1988]
Human intelligence develops situated in a multimodal environment [Gibbs, 2005].[ , ]The human mind makes use of multiple representations and problem-solving strategies [Fuster, 2003]. ]The brain consists of functional modules which are localized in subcortical areas but work togetheron the whole-brain scale [Grillner et o e w o e b sc e [G e eal., 2006]. Humans can integrate the multiple tasks into a coherent solution [Jones, 2004].2004]. Humans are versatile and come up with many new ideas and solutions to a given problem [Minsky, 2006].
Principles of Learning: Modern ConceptsPrinciples of Learning: Modern ConceptsPrinciples of Learning: Modern ConceptsPrinciples of Learning: Modern Concepts
Types of learning: Accretion, tuning, restructuring (e grestructuring (e.g., Rumelhart & Norman, 1976)Encoding specificityEncoding specificity principle (Tulving, 1970’s)Cellular and molecular basis of learning andbasis of learning and memory (Kandel et al., 1990’s) Conceptual blend andConceptual blend and chemical scramble (e.g., Feldman, 2006)
Principles of Information Processing in thePrinciples of Information Processing in thePrinciples of Information Processing in the Principles of Information Processing in the BrainBrain
The Principle of Uncertainty4Precision vs. predictionp
The Principle of Nonseparability “UN-IBM”4Processor vs. memoryy
The Principle of Infinity4Limited matter vs. unbounded memory
The Principle of “Big Numbers Count”4Hyperinteraction of 1011 neurons (or > 1017 molecules)
The Principle of “Matter Matters”4Material basis of “consciousness” [Zhang, 2005]
Cognitive Learning and MemoryCognitive Learning and Memory
Toward HumanToward Human--Level Machine Learning: Level Machine Learning: M l i d l M G (MMG)M l i d l M G (MMG)Multimodal Memory Game (MMG)Multimodal Memory Game (MMG)
But, I'm getting married tomorrowBut, I'm getting married tomorrowBut, I'm getting married tomorrowBut, I'm getting married tomorrowBut, I'm getting married tomorrowBut, I'm getting married tomorrowWell, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.
Well, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.But, I'm getting married tomorrowW ll b IBut, I'm getting married tomorrowW ll b IBut, I'm getting married tomorrowW ll b IBut, I'm getting married tomorrowW ll b I
Well, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.
Well, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.
Well, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.
Well, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.g g g pAre you thinking about me?But if you are, call me tonight.
g g g pAre you thinking about me?But if you are, call me tonight.
Well, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.A thi ki b t ?
Well, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.A thi ki b t ?
Well, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.A thi ki b t ?
Well, maybe I am...I keep thinking about you.And I'm wondering if we made a mistake giving up so fast.A thi ki b t ?
g g g pAre you thinking about me?But if you are, call me tonight.
g g g pAre you thinking about me?But if you are, call me tonight.
g g g pAre you thinking about me?But if you are, call me tonight.
g g g pAre you thinking about me?But if you are, call me tonight.
Are you thinking about me?But if you are, call me tonight.Are you thinking about me?But if you are, call me tonight.Are you thinking about me?But if you are, call me tonight.Are you thinking about me?But if you are, call me tonight.
Three ExperimentsThree ExperimentsThree ExperimentsThree Experiments
S G iSentence Generation4Learn: a linguistic recall memory from a sentence corpus4Given: a partial or corrupt sentencep p4Generate: a complete sentence
Image-to-Text Translation4 i j i d l f i i4Learn: an image-text joint model from an image-text pair corpus4Given: an image (scene)4Generate: a text (dialogue of the scene)( g )
Text-to-Image Translation4Learn: an image-text joint model from an image-text pair corpus4Gi (di l )4Given: a text (dialogue)4Generate: an image (scene of the dialogue)
Experiment 1: Learning Linguistic MemoryExperiment 1: Learning Linguistic MemoryExperiment 1: Learning Linguistic MemoryExperiment 1: Learning Linguistic Memory
D i f dDataset: scripts from dramas4Friends4HouseHouse4244Grey Anatomy 44Gilmore Girls 4Sex and the City
Training data: 289 468 sentencesTraining data: 289,468 sentences Test data: 700 sentences with blanksVocabulary size: 34 219 wordsVocabulary size: 34,219 words