Chess and machine intuition - George Atkinson, 1993.pdf

(8181 KB) Pobierz
CHESS
A
N
D
MACHI NE
INTUITION
GEORGE ATKINSON
44 0342748 0
CHESS AND MACHINE INTUITION
w ithdraw n
u n i v e
;:: ;
t y c
-
i r
o f
,c
s h ir e
LIBRARY. HATFIELD CAMPUS,
HATHS.D. AL10&AD
CONTROL
CLASS
15ViSb Q2
\ '7
COLLECT.Op
otvi-
. -«ATU3
ITEM
Copyright © 1993 by Ablex Corporation
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, microfilming, recording, or otherwise, without per­
mission of the publisher.
Printed in the United States of America
Library o f Congress Cataloging-in-Pubiication Data
Atkinson, George W.
Chess and machine intuition / George W. Atkinson.
p.
cm. — (Ablex series in artificial intelligence)
Includes bibliographical references and index.
ISBN 0-89391-901-2
1. Chess—Data processing. 2. Artificial intelligence.
I. Title. II. Series.
GV1449.3.A87 1993
794.1'72—dc20
93-28529
CIP
Ablex Publishing Corporation
355 Chestnut Street
Norwood, NJ 07648
Contents
Preface
1 Did Someone Say Ten Years?
An overview of the most conspicuous branch of artificial
intelligence, machine chess, in which a breakthrough is
always expected within a decade; a synopsis of the book.
vii
1
2 The Rise of Mechanical Automata
Von Kempelen’s 1769 “Turk” takes on all comers at
chess; Charles Babbage designs an Analytic Engine;
Torres y Quevedo builds an electro-mechanical
endgame machine.
15
3 B. P.
A group o f chess players and mathematicians at a secret
British wartime facility build machines to crack ciphers
and, for recreation, design chess machines.
23
4
Minimax
37
Assigning values to chess positions allows comparison
o f alternative branches o f a move tree; Claude Shannon
describes Alan Turing tests, and first programs run.
5 Brute Force
Amateur’s chess knowledge proves codifiable, the chess-
master’s does not; knowledge-based machines yield to
brute-force computation; computer tournaments become
a spectator sport.
53
V
vi
CONTENTS
6 Human Intuition
73
Humans are poor calculators, but exploit brain's pattern
recognition to play terrific chess; psychologists show
expertise is result o f trained intuition.
7 Human Versus Machine
85
John Henry beats the steam drill, David Levy conquers
CHESS 4.7, and Garry Kasparov outplays Deep Thought;
calculating amateurs lose while intuitive masters win.
8 Custom-Built Hardware
Chess machines on a chip and custom-built circuitry
amplify brute-force capability.
103
9 Computable Subgames
119
Human theoretical knowledge o f chess increases through
machine-assisted computation; exact endgame
computations provide unexpected results.
10 Machine Learning
131
It ain’t smart if it always makes the same mistakes;
some machines learn from experience; others induce rules
from examples, and can acquire intuitive knowledge.
11 Machine Intuition
149
Computing devices coupled to an environment mimic
neural systems for intuitive information processing and
offer hope for knowledge-based chess machinery.
Appendix A:
Appendix B:
Chess Notation
Torres y Quevedo’s Mating Algorithm
161
163
169
157
Appendix
C:
Recursive Programming and the
Minimax Algorithm
References
Author Index
Subject Index
171

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika:

Zgłoś jeśli naruszono regulamin