PANG.N I NG TAN Mich igan Sta te Un ivers i ty

MICHAEL STEINBACH Univers i ty o f M innesota

VI PI N KU MAR Univers i ty o f M innesota and Army High Performance Computing Research Center

+f.f_l crf.rfh. .W if f

aqtY 6l$

t .T.R.C.

i’&’ufe61ttt1/. Y t. $t,/,1′

n,5 . 7 V ’48!

Boston San Francisco NewYork

London Toronto Sydney Tokyo Singapore Madrid

MexicoCity Munich Paris CapeTown HongKong Montreal

G.R r+6,q

If you purchased this book within the United States or Canada you should be aware that it has been wrongfirlly imported without the approval of the Publishel or the Author.

T3 Loo 6

– {)gq* 3 AcquisitionsEditor Matt Goldstein

ProjectEditor Katherine Harutunian Production Supervisor Marilyn Lloyd Production Services Paul C. Anagnostopoulos of Windfall Software Marketing Manager Michelle Brown Copyeditor Kathy Smith Proofreader IenniferMcClain Technicallllustration GeorgeNichols Cover Design Supervisor Joyce Cosentino Wells Cover Design Night & Day Design Cover Image @ 2005 Rob Casey/Brand X pictures hepress and Manufacturing Caroline Fell Printer HamiltonPrinting

Access the latest information about Addison-Wesley titles from our iWorld Wide Web site: http : //www. aw-bc.com/computing

Many of the designations used by manufacturers and sellers to distiriguish their products are claimed as trademarks. where those designations appear in this book, and Addison- Wesley was aware of a trademark claim, the designations have been printed in initial caps or all caps.

The programs and applications presented in this book have been incl,[rded for their instructional value. They have been tested with care, but are not guatanteed for any particular purpose. The publisher does not offer any warranties or representations, nor does it accept any liabilities with respect to the programs or applications.

Copyright @ 2006 by Pearson Education, Inc.

For information on obtaining permission for use of material in this work, please submit a written request to Pearson Education, Inc., Rights and Contract Department, 75 Arlington Street, Suite 300, Boston, MA02II6 or fax your request to (617) g4g-j047.

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 any other media embodiments now known or hereafter to become known, without the prior written permission of the publisher. printed in the united States of America.

lsBN 0-321-42052-7

2 3 4 5 67 8 9 10-HAM-O8 07 06

our famili,es

Preface

Advances in data generation and collection are producing data sets of mas- sive size in commerce and a variety of scientific disciplines. Data warehouses store details of the sales and operations of businesses, Earth-orbiting satellites beam high-resolution images and sensor data back to Earth, and genomics ex- periments generate sequence, structural, and functional data for an increasing number of organisms. The ease with which data can now be gathered and stored has created a new attitude toward data analysis: Gather whatever data you can whenever and wherever possible. It has become an article of faith that the gathered data will have value, either for the purpose that initially motivated its collection or for purposes not yet envisioned.

The field of data mining grew out of the limitations of current data anal- ysis techniques in handling the challenges posedl by these new types of data sets. Data mining does not replace other areas of data analysis, but rather takes them as the foundation for much of its work. While some areas of data mining, such as association analysis, are unique to the field, other areas, such as clustering, classification, and anomaly detection, build upon a long history of work on these topics in other fields. Indeed, the willingness of data mining researchers to draw upon existing techniques has contributed to the strength and breadth of the field, as well as to its rapid growth.

Another strength of the field has been its emphasis on collaboration with researchers in other areas. The challenges of analyzing new types of data cannot be met by simply applying data analysis techniques in isolation from those who understand the data and the domain in which it resides. Often, skill in building multidisciplinary teams has been as responsible for the success of data mining projects as the creation of new and innovative algorithms. Just as, historically, many developments in statistics were driven by the needs of agriculture, industry, medicine, and business, rxrany of the developments in data mining are being driven by the needs of those same fields.In 5to 10pages (not includingtitle, Table of Contents, and References) explainthe virtues of the three clustering techniquesdiscussedin your assigned reading. The three techniques are K-Means, Agglomerative HierarchicalClustering, and DBSCAN. These are presented on page 533 of your assignedtext