tty10_15_pptx - Business and Computer Science

cabbagecommitteeAI and Robotics

Oct 24, 2013 (4 years and 2 months ago)

84 views

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Digital Planet
:

Tomorrow’s Technology


and You

George Beekman


Ben Beekman


Tenth Edition

Digital Planet:

Tomorrow’s Technology and You


Chapter 15

Is Artificial Intelligence Real?

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Chapter 15 Objectives


Explain the two basic approaches of artificial
intelligence research


Describe several hard problems that artificial
intelligence research has not yet been able to solve


Describe several practical applications of artificial
intelligence


Explain what robots are and give several examples
illustrating what they can

and can’t

do


Speculate about how our world might change as
artificial intelligence technology progresses

3

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Thinking About Thinking Machines


Definitions of
intelligence

include:


Ability to learn from experience


Power of thought


Ability to reason


Ability to perceive relations


Power of insight


Ability to use tools


Intuition

4

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Can Machines Think?


In 1950 Alan Turing proposed an “imitation game.”


Turing test:
involves two people and a computer


One person, the interrogator, sits alone in a room and
types questions into a computer terminal.


As answers to questions appear, the interrogator
attempts to guess whether those answers were typed
by a person or by the computer.


By repeatedly fooling interrogators, a computer can
demonstrate intelligent behavior.

5

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Can Machines Think?

6

The Turing
test

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

What is Artificial Intelligence?

Artificial intelligence is the study of ideas that enable computers to do the
things
that make people seem intelligent.


Patrick Henry Winston
, in Artificial Intelligence

Artificial intelligence is the study of how computers do things at which,
at the
moment, people are better.


Elaine Rich
, in Artificial Intelligence

Artificial intelligence is the study of the computations that make it possible to
perceive, reason, and act.


Patrick Henry Winston
, in Artificial Intelligence

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Opening Games


Checkerboard: first popular domain for AI research


Chess and checkers had clearly defined rules and
unmistakable goals.


Searching:
looking ahead at the possibilities


Heuristics:
rule of thumb


Pattern recognition:
recognizing recurring patterns


Machine learning:
learning from experience

8

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Natural
-
Language Communication


In Turing’s test of intelligence, a computer is
considered intelligent if it can pass as a person in a
typed conversation.


Many problems relate to recognizing and
reproducing human speech.


Natural
-
language text poses significant software
challenges.

9

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Machine Translation Traps


Automatic translation:
offered hope for increased
communication between scientists during the tense
Cold War years


Parsing program:
would analyze sentence structure
and another program would look up each word in a
translation dictionary and substitute the appropriate
word


This word
-
by
-
word approach failed.

10

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Conversation without Communication


ELIZA was one of the first programs to converse in a
limited form of natural language.


Designed to simulate role of a therapist in a typed
conversation with a patient.


A session could deteriorate into nonsense dialogue
laced with grammatical errors and inappropriate
responses.

11

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Nonsense and Common Sense


A major part of the problem with natural language
communication is massive vocabulary.


English contains hundreds of thousands of words,
many of which have multiple meanings.


Computers are more successful dealing with natural
-
language
syntax

a set of rules for constructing
sentences from words

than with
semantics

the
underlying meaning of words and phrases.

12

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Knowledge Bases and Expert Systems


The human brain excels at manipulating knowledge.


Computers are better at handling data than
knowledge.


AI researchers have developed, and continue to
develop, techniques for representing knowledge in
computers.

13

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Knowledge Bases


A database contains only facts.


Knowledge base:
contains a system of rules for
determining and changing the relationship among
those facts


Ideas stored in a knowledge base can be reorganized
as new information changes their relationships.

14

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Artificial Experts


Expert: someone with an extraordinary amount of
knowledge within a narrow domain


Expert system:
software program designed to
replicate the decision
-
making process of a human
expert


At the foundation of every expert system is a
knowledge base representing ideas from a specific
field of expertise.

15

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Artificial Experts (cont.)


Expert systems derive their knowledge from experts.


Systems that draw on other sources are called
knowledge
-
based systems.


A knowledge base commonly represents knowledge
in the form of if
-
then rules, such as these:


If the engine will not turn over and the lights do not
work, then check the battery.


If checking the battery shows it is not dead, then
check the battery connectors.

16

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Expert Systems in Action


Some of the first successful expert systems were
developed around medical knowledge bases.


The MYCIN medical expert system outperformed
many human experts in diagnosing diseases.


The business community has accepted the use
of
expert
systems more readily than the medical
community.

17

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Expert Systems in Action (cont.)


Examples of Expert Systems


Microsoft Windows Help software provides advice,
suggestions, and solutions for common problems and
errors.


American Express uses an expert system to automate
the process of checking for fraud and misuses of its
no
-
limit credit card.


At Blue Cross/Blue Shield of Virginia, an expert system
automates insurance claim processing.

18

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Expert Systems in Perspective


An expert system can:


Help train new employees


Reduce the number of human errors in complex tasks


Take care of routine tasks


Provide expertise when no experts are available


Preserve the knowledge of experts after they leave


Combine the knowledge of several experts


Make knowledge available to more people

19

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Pattern Recognition:

Making Sense of the World


Pattern recognition:
involves identifying recurring
patterns in input data to understand or categorize
that input


Pattern recognition programs represent half of the AI
industry.

20

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Pattern Recognition:

Making Sense of the World (cont.)


Face identification


Fingerprint
identification


Handwriting recognition


Optical character
recognition


Automatic voice
recognition


21


Biological slide analysis


Surveillance satellite
data analysis


Weather forecasting


Robot vision


Expert systems


Scientific data analysis


Pattern Recognition Applications

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Image Analysis


Image analysis:
process of identifying objects and
shapes in a photograph, drawing, video, or other
visual image


Used for everything from autofocusing cameras on
human faces to piloting cruise missiles


Today’s computers are capable of running image
-
processing software with practical applications.


Security programs enable PCs with video cameras to
recognize faces of users with high degree of reliability.


22

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Optical Character Recognition


Optical character recognition:
not perfect


The first step in general OCR is to scan the image of
the page into the computer’s memory with a
scanner, digital camera, or fax modem.


Before a computer can process the text on a page, it
must recognize the individual characters and convert
them to text codes.

23

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Optical Character Recognition (cont.)


Optical character recognition (OCR) software:
locates and identifies printed characters embedded
in images


The process of recognizing text in a variety of fonts
and styles is difficult for machines.

24

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Optical Character Recognition (cont.)


OCR programs use several techniques, including:



Segmentation of the page into pictures, text blocks,
and (eventually) individual characters



Scaled
-
down expert system technology for
recognizing underlying rules that distinguish letters


Context “experts” to help identify ambiguous letters
by their context


Learning from actual examples and feedback from a
human trainer

25

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Automatic Speech Recognition

26


Automatic
speech recognition
systems use pattern recognition
techniques, including these:


Segmentation of input sound
patterns into individual words
and phonemes


Expert rules for interpreting
sounds


Context “experts” for dealing with
ambiguous sounds


Learning from a human trainer


Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Talking Computers


It’s easier for machines to speak passable English or
Chinese than to recognize it.


There are many applications for voice output
including:


Preschool education


T
elephone communication


N
avigation guidance systems in cars


R
eading machines for users who are visually impaired

27

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Talking Computers (cont.)


Many computer applications speak as humans by
playing prerecorded
digitized speech
(along with
other
digitized sounds
) stored in memory or on disk.


Digitized speech is practical and reliable for:


An application with a limited vocabulary (reciting
telephone numbers for automated directory
assistance)


An application with limited choices (an interactive
educational game with short prerecorded speeches)


28

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Talking Computers (cont.)


Recorded speech doesn’t work for applications in
which the text to be spoken is unpredictable.


These applications require
text
-
to
-
speech

conversion

the creation of
synthetic speech
by
converting text files into phonetic sounds.


With
speech synthesis
software or hardware, PCs can
recite anything you can type, but with voices that
sound artificial and robotic.

29

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Neural Networks


Neural networks

(or
neural nets
): distributed,
parallel computing systems inspired by the structure
of the human brain


A neural network uses a network of a few thousand
simpler processors called neurons.


Researchers hope that neural networks may
someday provide hearing for the deaf and eyesight
for the blind.

30

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Neural Networks (cont.)


Neural net must go through a
series of trials in which circuit
patterns that produce
incorrect guesses are
weakened


Patterns that produce correct
guesses are strengthened


The result is a circuit pattern
that can recognize the letter A
in a variety of forms

31

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Question
-
Answering Machines


Question
-
answering machine:
should be able to
understand natural
-
language questions and provide
answers to those questions by drawing on a base of
stored knowledge from a broad array of disciplines


Many applications and Web sites have been able to
answer questions, but only if they were carefully
phrased and limited to specific subjects.


They were not able to find answers not stored in
their databases.

32

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Question
-
Answering Machines (cont.)


IBM’s Watson is probably the most ambitious
question
-
answering machine to date.


Watson pushes the limits of artificial intelligence in
several key areas:


Natural
-
language processing


Knowledge bases


Pattern recognition


33

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

The Robot Revolution


AI technologies are most visible in the field of
robotics.


Vision, hearing, pattern recognition, knowledge
engineering, expert decision making, natural
-
language understanding, speech

they all come
together in today’s robots.

34

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

What Is a Robot?


Robots:

similar to other kinds of computer
technology people use every day


A robot is a computer
-
controlled machine designed
to perform specific manual tasks.


A robot sends commands to joints, arms, and other
moving parts.


Most modern robots include input
sensors

that
enable them to correct or modify actions based on
feedback.

35

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

What Is a Robot? (cont.)

36

A Robot Is a Computer

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Steel
-
Collar Workers


Robots offer several advantages:


Many robots save labor costs and can work 24 hours a
day, 365 days a year without vacations or strikes.


Robots
also can improve
quality and increase
productivity.


Robots are ideal for jobs such as cleaning up
hazardous waste and salvaging undersea wreckage
from downed planes.

37

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

AI Implications and Ethical Questions


Even when they don’t work very well, AI programs
generate emotional responses in people who use them.


As it matures, AI technology finds its way out of the
research lab and into the marketplace.


AI grows as networked systems learn from each other,
correct each others’ work, and provide “insights” that
come from different perspectives.

38

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

AI Implications and Ethical Questions
(cont.)



39

Some think artificial intelligence
is the natural culmination of the
evolutionary process

that the
next intelligent life
-
form on Earth
will be based on silicon rather
than the carbon that is the basis
of human life.

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Artificial Life


Many recent successes in artificial intelligence have
come from a related field known as
artificial life

(or
Alife
).


Artificial life is the study and creation of systems
related to life, its processes, and its evolution.


Artificial life comes in three basic forms: soft Alife,
hard Alife, and wet Alife.

40

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Chapter 15 Summary


Most AI research has focused on making computers
do things at which people generally are better.


AI programs employ a variety of techniques,
including searching, heuristics, pattern recognition,
and machine learning to achieve their goals.


Natural
-
language communication is one of the most
important areas of AI study.

41

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Summary (cont.)


A knowledge base contains facts and a system for
determining and changing the relationships among
those facts.


Expert systems are programs designed to replicate
the decision
-
making process of human experts.


Pattern recognition is an important area of AI
research that involves identifying recurring patterns
in input data.


42

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

Summary (cont.)


Question
-
answering systems combine natural
-
language processing, knowledge bases, pattern
recognition, and other AI technologies.


A robot is a computer
-
controlled machine designed
to perform specific manual tasks.


Artificial life is the study and creation of systems
related to life, its processes, and its evolution.


43

Copyright © 2012 Pearson Education, Inc. publishing as Prentice Hall

All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any
means, electronic, mechanical, photocopying, recording, or
otherwise, without the prior written permission of the publisher.
Printed in the United States of America.

44