DATA MINING TECHNIQUES eet yor" Arun K Pujari seonrte seo a ETE WITHIN wy Universities Press smadAnt Aes FILLER. Whoa Universes Press nia) Private Reginve ier BETEEUA 58 1841 Timayatonge pera 500029 (8.75. lie lin ance cron 2 Univerennas Poss indian Patel nated 200 Fis sasched2081 es ines 2008 ISBS-L5 97H RI TS71 ARC SSeS.) RUPE Od pared Hysbrshea 1039 Pete tot se Graphs Pritts Hyleasad $00 013 Pati i aiversitie Pe edit Pvt Limited SE MMUA 26.7041 Himayamaeae pra SHIIES CA, Ea w ay ronthor Late Sd. Tatas Des = aver: fragile and weak tad se ws the pillar of strength Jor all her children ‘nd er husband fa team on ConTENTS AT A GLANCE Iwrropuerion ‘Data WAREHOUSING Dara Minin: Associrion RULES Cuustenine TecHmoves Decision Tree TecHuraues ‘OtHen TecumauEs Wee Minis ‘Temponat ano Srarian Data MINING List oF Contents Fomor “i Protoatx Prairace ACKRENLEDCEMENT 1 tvexopucras ' LA Iseoduetion 1 12 ata Ming as & Subject, 4 13 Gide tothis Book 5 2 Dara Wanenowsse 7 21 rwraduetion 7 22 What isa Data Warehuuse? ° 23 Definition lo 2.4 Multidimensiousl Date Medet W 2S OLAP Operations u 26 Warehouse Schema “0 2.7 Dota Warehousing Arstitootore cy 2B Warchnuse Server a 29 Metwiata % 2.10 OLAP Engine 2B BLL Data Warchouse Backend Process cf 2.12 Other Feacuss 31 213 Summary x Exercioos 3 Bibliography 0 3 Dara Moms 2 3 lateedvetion 2 32 Whalis Date Mining? B 3.3 Data Mining: Detiutions 4 34 KDD vs. Paw Mining a 34 DBMS wsDM ” 36 Other Relued Areas st 32 DM Techniques 31 341 Other Mining Prablems 3 wii Des Mining Tenis $9 Teaver and Caallanges in BM. 8 BAN DM Apploation Areas “0 BAL DM Applicetions—Cuve Sues a $12 Conclusion 6 Furs Realing KG Exercises o Bibliowsehy co 4 Awpuauias Russ oo 4. Introduction ° 42. Whatsgan Association Rule? i 4.3. Mottedste Dicenvar Assoo-ation Rules B 44° APeiur Algocithay 45 Partition Algorithm 4.6 Fincer-Ssare: Algorithm BS 4.7 Dytuie lentset Coursing Algorithm 0 44 Petree Cromth Alport, on 49. Digeus.on or iffeenl Algoritims 8 440. feorementl Auris 10 SLL. Roger Algerithn 102 4.12 Genecilized Association Rule ns 43. Assocation Rules With Hem Cone 414: Summary Facet Reading Cacrcises Riliography § Ciustenye Teen wes Tnlroduction Chasen Paraligats Paririoning Slgorithns keMedoid Aigorshms CLARA CLARANS, Teeearhica: Clustoring DESCAN BIRCH CURE, Categorical Case AlgorStuns SUIRK ROCK CACTES, 515 Conchuon Fuchor Reading Exotics Bitliography Dosen Deets 6.1 Traraducsion 6.2 What isa Dovsion Te 63 Tree Construction Principle fia Best Spht 65. Splitang indices 66 Splicing Crvera 62 Devision Tioe Constracton Aluorithrs fd CART 6.13. Devssion Troe Construction vita resorting 6.14 Rainforest 6.13 Approeimate Methods 6.16 CLOUDS 617 BOAT 6.18 Proning Technique 6.19 ategeatin of poning and consiroction ‘Summary: An Edel Algosivhm Otier Topics 622 Conclusion other Reulimg Hrcreies Bibliography Onuen Tecngens TL Introduction 72. What is a Nowal Netwusk? 72 Learning ia NM 74 Unvarervisel Leaning 15. Data Mining using NN: A Case Seedy 7.6 Genetic gorithm 73 Rouph Sets ER Suppotl Veotur Machines 79° Conclusion on 1a ra 1a 1 19 Ge Mining Tedsigues ure Reading Excrcies Tibliegraphy 8 Wee Minin SL Inledust: 2) WebMining ¥3- Web Content Mining A WebStrucore Mii, sage Mir ke Ten Mine Cesmuctatse Tex 83. Episorte Rule Duscovery fur Texts K9-Hieraschy vf Catogories 8.19 TerClusering 14 Corclusive Furker Resding Bitioazaphy Addendum 9 Tresrogas asp Seanad Daa MINING 9. taleedoction 92 Weat is Texapoeal Bala Mining? 9.3 Temporal Assuciatan Raise 9A Sequence Minty 93 The GSP Algorithes 6 SPADE 97 SPT 93 WUM 9.9 Epil Discovery 9.10 sent Predieson Pendle 941 Time sores Analysis AE Spall Mince 3.13 Spanal Mimsy Tasks 944 Spatial Chute, 95. Spatia! Freud 3.16 Cypelusion urtaer Reading Exerises Bitliowsephy INU FOREWORD ‘Vasc amouns of peratvnal Ua ae routinely coMlected and sore ave inte archives of many orguncanins. Ta cake 8 sinple example, the ralway reservation syn haz beet vperiona! for over a deca and a vast srivuat ul Usa f generated each day on tai bookings. Much of this data is probably archived for audi pugpoven. "hi archived coperatianaldsta can be efecively used fer tached! stratsie mitayennent ofthe raya. ue insince, hy analyzing the -eservaion data wt auld he possible fd oul wae paler wr various sectors and wae ith add or iemowe bute i crtin tas, to dtide om the mis of varivus elases uf accommodation, ete, This, Inwever, was not Feasible und ‘oeenmy duc w limitations in beth hardware and software. Tote lt Ave yeas there ba boon a trontendous improversea i hardware 5:2 MBE af main memory, 40 GB ask and CPU clock vf £ 6 GAY as now avalb'e an a PC Thus, cemmputerprngeans which sil massive arscums uf opecaional dul, eecogniza patterns and previde ints 12 fornia Irytheses for acti aud static Ueision-making can naw be executed 2 easerable ‘me. This has aponed op file ate of rewazch ro frmulae ayn algocitms Ro mining archival data i forrulce and Wal hypotkece. Aa uray of Has Fm computer stience. sates and manageron scones ae beg pied in Us are, This Suh: is cme active research aa, 1 ars extemely tuppy tbat Profeswor Aran K, Puja bat writen this book on Mack ‘Mining Techaigues Il ley adition wo the exiting comple sence iterate and ‘will be a boon to studeals Mrofeeor Pyar. has been an aclveratearchey i thi area, He Jas als intrdooed a coorss on Dats Ming a tke University of Hydcabad and ns bora feaching ic His rice expeconoe in teaching amt research is evident in the aeleeion of topics and tele eatment inthis hook. The book sis wilh brie intoduction followed bya dctafed chap on dala naccheusing,Ithas chapters whisk disease a Tange number of sivereealgorigzms using saucalion mle, clustering sechniqus, dsicion tre techniques, rneuual newark, genetics, rough sel, ad supper weer machines. A cel aspect ofthis nok iss diseussion ofthe enecying are of wel ming of extual dat, A Tage Nucuber freferences are cit and the mestmet slid and pw date, Irecazimend this Pook 2 sludens wha want io explre tis ipeisul sect V-Rajaramen, Manarry Professor Super Compuler Edueaiion & Research Cer Tadign Ise of Scien, Bangalore