Overview of Artificial Intelligence.ppt by pptfiles

VIEWS: 7 PAGES: 27

									Overview of
Artificial Intelligence
   Slides from Jean-Claude Latombe, Kai Yu,
   Sebastian Thrun, Peter Norvig, and etc.
    Jianlin Feng
    School of Software
    SUN YAT-SEN UNIVERSITY
    What is Intelligence?

n   Intelligence: from Webster online
    1) the ability to learn or understand or to deal with
        new or trying situations.
    2) the ability to apply knowledge to manipulate one's
        environment or to think abstractly as measured by
        objective criteria (as tests)
n   in particular,
    q    the ability to solve novel problems
    q   the ability to act rationally
    q   the ability to act like humans
    What is AI -- found on the Web …
n   AI is the simulation of intelligent human processes
         Intelligent
n   AI is the reproduction of the methods or results of human
           behavior
    reasoning or intuition
                                          Computer
n   AI is the study of mental faculties through the use
    computational methods

n   Using computational models to simulate intelligent
    behavior
                       Humans
n   Machines to emulate humans
What is AI?
(Russell&Norvig)

Discipline that systematizes and automates
reasoning processes to create computer
systems that:
     Act like humans       Act rationally
    Think like humans     Think rationally




                                             4
          Act like humans            Act rationally
         Think like humans          Think rationally

§   The goal of AI is to create computer systems that
    perform tasks regarded as requiring intelligence when
    done by humans
§   à AI Methodology: Take a task at which people are
    better, e.g.:
    •   Prove a theorem
    •   Play chess
    •   Plan a surgical operation
    •   Diagnose a disease
    •   Navigate in a building
    and build a computer system that does it automatically

§   But do we want to duplicate human imperfections?
                                                             5
     Act like humans            Act rationally
    Think like humans          Think rationally


§   Here, how the computer performs tasks does
    matter
§   The reasoning steps are important
§   à Ability to create and manipulate symbolic
    knowledge (definitions, concepts, theorems, …)
§   What is the impact of hardware on low-level
    reasoning, e.g., to go from signals to symbols?


                                                      6
      Act like humans            Act rationally
     Think like humans          Think rationally


§   Now, the goal is to build agents that always make the
    “best” decision given what is available (knowledge,
    time, resources)
§   “Best” means maximizing the expected value of a
    utility function
§   à Connections to economics and control theory
§   What is the impact of self-consciousness, emotions,
    desires, love for music, fear of dying, etc ... on human
    intelligence?


                                                               7
    Can Machines Act/Think
    Intelligently?
Turing Test:
§   http://plato.stanford.edu/entries/turing-test/
§   Test proposed by Alan Turing in 1950
§   The computer is asked questions by a human
    interrogator. It passes the test if the
    interrogator cannot tell whether the responses
    come from a person
§   Required capabilities: natural language
    processing, knowledge representation,
    automated reasoning, learning,...
§   No physical interaction



                                                     8
An Application of the Turing Test

§   CAPTCHA:
    §   Completely Automatic Public Turing tests to
        tell Computers and Humans Apart
§   E.g.:
    •   Display visually distorted words
    •   Ask user to recognize these words
§   Example of application: have only humans
    open email accounts


                                                      9
    Can Machines Act/Think
    Intelligently?
§   Yes, if intelligence is narrowly defined as
    information processing
    AI has made impressive achievements showing that
    tasks initially assumed to require intelligence can be
    automated
    But each success of AI seems to push further the limits
    of what we consider “intelligence”




                                                        10
    Can Machines Act/Think
    Intelligently?
§   Yes, if intelligence is narrowly defined as
    information processing
    AI has made impressive achievements showing that
    tasks initially assumed to require intelligence can be
    automated
§   Maybe yes, maybe not, if intelligence is not
    separated from the rest of “being human”



                                                             11
    Some Big Open Questions

§   AI (especially, the “rational agent” approach) assumes
    that intelligent behaviors are only based on information
    processing? Is this a valid assumption?
§   If yes, can the human brain machinery solve problems
    that are inherently intractable for computers?
§   In a human being, where is the interface between
    “intelligence” and the rest of “human nature”, e.g.:
    •   How does intelligence relate to emotions felt?
    •   What does it mean for a human to “feel” that he/she
        understands something?
§   Is this interface critical to intelligence? Can there
    exist a general theory of intelligence independent of
    human beings? What is the role of the human body?

                                                              12
    Some Big Open Questions

§   AI (especially, the “rational agent” approach) assumes
    that intelligent behaviors are based on information
     In the movie I, valid assumption? impressive
    processing? Is this a Robot, the most impressive
     In the movie I, Robot, the most
§    feature the the robots is not their ability to
    If yes, can of the robots machinery solveability to
     feature of human brain is not their problems
     solve complex problems, for how they blend
    that arecomplex problems, but computers? blend
     solve inherently intractable but how they
§    human-like reasoning with other between
    In a human being, where is the interface key
      human-like reasoning with other key
    “intelligence” and the rest of “human nature”, e.g.:
     aspects of human beings (especially, self-
                 of human beings (especially, self-
      aspects intelligence relate to emotions felt?
       How does
     consciousness, fear of dying, distinction
    §

       What does it mean fear of dying, distinction
    § consciousness,for a human to “feel” that he/she

     between right and wrong)
       understands something?
      between right and wrong)
§   Is this interface critical to intelligence? Can there
    exist a general theory of intelligence independent of
    human beings? What is the role of the human body?

                                                            13
    Main Areas of AI

§   Knowledge representation
    (including formal logic)      Agent                 Perception
§   Search, especially                       Robotics
    heuristic search (puzzles,
    games)                       Reasoning
§   Planning                                 Search
§   Reasoning under                                       Learning
    uncertainty, including
    probabilistic reasoning               Knowledge Constraint
                                             rep.
§   Learning                     Planning          satisfaction
§   Agent architectures
§   Robotics and perception
§   Natural language             Natural
                                                ...       Expert
    processing                   language
                                                         Systems
                                                                     14
 A (Short) History of AI




1/25/2014          15
Predictions and Reality … (1/3)

n   In the 60’s, a famous AI professor from MIT
    said: “At the end of the summer, we will have
    developed an electronic eye”
n   As of 2002 2011, there is still no general
    computer vision system capable of
    understanding complex dynamic scenes
n   But computer systems routinely perform road
    traffic monitoring, facial recognition, some
    medical image analysis, part inspection,etc.
Predictions and Reality … (2/3)


n   In 1958, Herbert Simon (CMU) predicted
    that within 10 years a computer would be
    Chess champion
n   This prediction became true in 1998
n   Today, computers have won over world
    champions in several games, including
    Checkers, Othello, and Chess, but still do
    not do well in Go
Predictions and Reality … (3/3)



n   In the 70’s, many believed that computer-controlled
    robots would soon be everywhere from manufacturing
    plants to home
n   Today, some industries (automobile, electronics) are
    highly robotized, but home robots are still a thing of
    the future
n   But robots have rolled on Mars, others are performing
    brain and heart surgery, and humanoid robots are
    operational and available for rent (see:
    http://world.honda.com/news/2001/c011112.html)
Success Stories
n   Deep Blue defeated the reigning world chess champion Garry
    Kasparov in 1997

n   AI program proved a mathematical conjecture (Robbins conjecture)
    unsolved for decades

n   NASA's on-board autonomous planning program controlled the
    scheduling of operations for a spacecraft

n   Proverb solves crossword puzzles better than most humans

n   Robot driving: DARPA grand challenge 2003-2007

n   2006: face detection software available in consumer cameras

n   IBM’s QA system Watson beats human experts in Jeopardy! show
    Example: DARPA Grand Challenge
n   Grand Challenge
    q Cash prizes ($1 to $2 million) offered to first robots to complete a long
       course completely unassisted
    q Stimulates research in vision, robotics, planning, machine learning,
       reasoning, etc

n   2004 Grand Challenge:
    q 150 mile route in Nevada desert

    q Furthest any robot went was about 7 miles

    q … but hardest terrain was at the beginning of the course



n   2005 Grand Challenge:
    q 132 mile race

    q Narrow tunnels, winding mountain passes, etc
                st       nd
    q Stanford 1 , CMU 2 , both finished in about 6 hours



n   2007 Urban Grand Challenge
    q Victorville, California
       Stanley Robot
Stanford Racing Team   www.stanfordracing.org




                  Next few slides courtesy of Prof.
                  Sebastian Thrun, Stanford University
SENSOR INTERFACE                            PERCEPTION                                          PLANNING&CONTROL                                   USER INTERFACE

 RDDF database       corridor
                                                                                                               Top level control                            Touch screen UI

                                                                                                              pause/disable command
                                                                                                                                                            Wireless E-Stop
 Laser 1 interface
                                           RDDF corridor (smoothed and original)                                 driving mode
 Laser 2 interface

 Laser 3 interface                                                           road center
                                                 Road finder                                                     Path planner
 Laser 4 interface                                  laser map


                                                                                    map                             trajectory
 Laser 5 interface                             Laser mapper                                                                                                  VEHICLE
 Camera interface                             Vision mapper
                                                                            vision map                                                                     INTERFACE
                                                                             obstacle list                     Steering control
 Radar interface                               Radar mapper
                                           vehicle state (pose, velocity)                                                                                  Touareg interface
                                                                                      vehicle
                                                                                       state                 Throttle/brake control
  GPS position                            UKF Pose estimation
                                                                                                                                                         Power server interface
                                           vehicle state (pose, velocity)
  GPS compass

  IMU interface                            Surface assessment
                                                                              velocity limit


  Wheel velocity

  Brake/steering
                           heart beats                   Linux processes start/stop                                          emergency stop

                                                                                             health status
                                         Process controller                                                     Health monitor
                                                                                                                                          power on/off
                                                                             data


   GLOBAL                                                             Data logger                                                                             File system
  SERVICES
                                                     Communication requests               Communication channels                                                  clocks



                                              Inter-process communication (IPC) server                                                                        Time server
Planning = Rolling out Trajectories
2004: Barstow, CA, to Primm, NV




               150 mile off-road robot race
                150 mile off-road robot race
               across the Mojave desert
                across the Mojave desert
               Natural and manmade hazards
                Natural and manmade hazards
               No driver, no remote control
                No driver, no remote control
               No dynamic passing
                No dynamic passing
               Fastest vehicle wins the race
                Fastest vehicle wins the race
               (and 22million dollar prize)
                (and million dollar prize)
Intelligent Systems in Your Everyday Life
n   Post Office
    q automatic address recognition and sorting of mail




n   Customer Service
    q automatic voice recognition



n   The Web
    q Identifying your age, gender, location, from your Web surfing

    q Automated fraud detection



n   Digital Cameras
    q Automated face detection and focusing



n   Computer Games
    q Intelligent characters/agents
Why is AI Hard?

n   Simple syntactic manipulation is not enough
Why is AI Hard?

    Computational intractability

•AI goal defined before notion of NP-completeness
•people thought to solve larger problems we
simply need larger/faster computers
•didn’t understand the notion of exponential
growth

								
To top