This book will give the reader a perspective into the core theory and practice of data mining and knowledge discovery (DM&KD). Its chapters combine many theoretical foundations for various DM&KD methods, and they present a rich array of examples—many of which are drawn from real-life applications. Most of the theoretical developments discussed are accompanied by an extensive empirical analysis, which should give the reader both a deep theoretical and practical insight into the subjects covered. The book presents the combined research experiences of its 40 authors gathered during a long search in gleaning new knowledge from data. The last page of each chapter has a brief biographical statement of its contributors, who are world-renowned experts. __Audience__ The intended audience for this book includes graduate students studying data mining who have some background in mathematical logic and discrete optimization, as well as researchers and practitioners in the same area. Cover 1 TABLE OF CONTENTS 6 FOREWORD 34 Chapter 1 ^A COMMON LOGIC APPROACH TOMINING AND PATTERN RECOGNITIONDATA 37 Chapter 2 *THE ONE CLAUSE AT A TIME (OCAT)APPROACH TO DATA MINING ANDKNOWLEDGE DISCOVERY 80 Chapter 3 *AN INCREMENTAL LEARNING ALGORITHMFOR INFERRING LOGICAL RULES FROMEXAMPLES IN THE FRAMEWORK OF THECOMMON REASONING PROCESS 123 Chapter 4 ^DISCOVERING RULES THAT GOVERNMONOTONE PHENOMENA 182 Chapter 5 ^LEARNING LOGIC FORMULAS ANDRELATED ERROR DISTRIBUTIONS 226 Chapter 6 ^FEATURE SELECTIONFOR DATA MINING 260 Chapter 7 ^TRANSFORMATION OF RATIONAL DATAAND SET DATA TO LOGIC DATA 286 Chapter 8 ^DATA FARMING: CONCEPTS AND METHODS 312 Chapter 9 ^RULE INDUCTION THROUGH DISCRETESUPPORT VECTOR DECISION TREES 338 Chapter 10 ^MULTI-ATTRIBUTE DECISION TREES ANDDECISION RULES 360 Chapter 11 *KNOWLEDGE ACQUISITION ANDUNCERTAINTY IN FAULT DIAGNOSIS:A ROUGH SETS PERSPECTIVE 392 Chapter 12 ^DISCOVERING KNOWLEDGE NUGGETS WITHA GENETIC ALGORITHM 428 Chapter 13'DIVERSITY MECHANISMS IN PITT-STYLEEVOLUTIONARY CLASSIFIER SYSTEMS 466 Chapter 14 ^FUZZY LOGIC IN DISCOVERINGASSOCIATION RULES: AN OVERVIEW 491 Chapter 15'MINING HUMAN INTERPRETABLEKNOWLEDGE WITH FUZZY MODELINGMETHODS: AN OVERVIEW 526 Chapter 16'DATA MINING FROM MULTIMEDIA PATIENTRECORDS 582 Chapter 17 ^LEARNING TO FIND CONTEXTBASED SPELLING ERRORS 627 Chapter 18 ^INDUCTION AND INFERENCEWITH FUZZY RULES FOR TEXTUALINFORMATION RETRIEVAL 658 Chapter 19 *STATISTICAL RULE INDUCTION IN THEPRESENCE OF PRIOR INFORMATION: THEBAYESIAN RECORD LINKAGE PROBLEM 683 Chapter 20 •SOME FUTURE TRENDS IN DATA MINING 723 SUBJECT INDEX 745 AUTHOR INDEX 755 A Common Logic Approach to Data Mining and Pattern Recognition....Pages 1-43 The One Clause at a Time (OCAT) Approach to Data Mining and Knowledge Discovery....Pages 45-87 An Incremental Learning Algorithm for Inferring Logical Rules from Examples in the Framework of the Common Reasoning Process....Pages 89-147 Discovering Rules That Govern Monotone Phenomena....Pages 149-192 Learning Logic Formulas and Related Error Distributions....Pages 193-226 Feature Selection for Data Mining....Pages 227-252 Transformation of Rational Data and Set Data to Logic Data....Pages 253-278 Data Farming: Concepts and Methods....Pages 279-304 Rule Induction Through Discrete Support Vector Decision Trees....Pages 305-326 Multi-Attribute Decision Trees and Decision Rules....Pages 327-358 Knowledge Acquisition and Uncertainty in Fault Diagnosis: A Rough Sets Perspective....Pages 359-394 Discovering Knowledge Nuggets with a Genetic Algorithm....Pages 395-432 Diversity Mechanisms in Pitt-Style Evolutionary Classifier Systems....Pages 433-457 Fuzzy Logic in Discovering Association Rules: An Overview....Pages 459-493 Mining Human Interpretable Knowledge with Fuzzy Modeling Methods: An Overview....Pages 495-550 Data Mining from Multimedia Patient Records....Pages 551-595 Learning to Find Context Based Spelling Errors....Pages 597-627 Induction and Inference with Fuzzy Rules for Textual Information Retrieval....Pages 629-653 Statistical Rule Induction in the Presence of Prior Information: The Bayesian Record Linkage Problem....Pages 655-694 Some Future Trends in Data Mining....Pages 695-716 2. Some Background Information 49 3. Definitions and Terminology 52 4. The One Clause at a Time (OCAT) Approach 54 4. 1 Data Binarization 54 4. 2 The One Clause at a Time (OCAT) Concept 58 4. 3 A Branch-and-Bound Approach for Inferring Clauses 59 4. 4 Inference of the Clauses for the Illustrative Example 62 4. 5 A Polynomial Time Heuristic for Inferring Clauses 65 5. A Guided Learning Approach 70 6. The Rejectability Graph of Two Collections of Examples 72 6. 1 The Definition of the Rej ectability Graph 72 6. 2 Properties of the Rejectability Graph 74 6. 3 On the Minimum Clique Cover of the Rej ectability Graph 76 7. Problem Decomposition 77 7. 1 Connected Components 77 7. 2 Clique Cover 78 8. An Example of Using the Rejectability Graph 79 9. Conclusions 82 References 83 Author's Biographical Statement 87 Chapter 3 AN INCREMENTAL LEARNING ALGORITHM FOR INFERRING LOGICAL RULES FROM EXAMPLES IN THE FRAMEWORK OF THE COMMON REASONING PROCESS, by X. Naidenova 89 1. Introduction 90 2. A Model of Rule-Based Logical Inference 96 2. 1 Rules Acquired from Experts or Rules of the First Type 97 2. 2 Structure of the Knowledge Base 98 2. 3 Reasoning Operations for Using Logical Rules of the First Type 100 2. 4 An Example of the Reasoning Process 102 3. Inductive Inference of Implicative Rules From Examples 103 3.
This book will give the reader a perspective into the core theory and practice of data mining and knowledge discovery (DM&KD). Its chapters combine many theoretical foundations for various DM&KD methods, and they present a rich array of examples—many of which are drawn from real-life applications. Most of the theoretical developments discussed are accompanied by an extensive empirical analysis, which should give the reader both a deep theoretical and practical insight into the subjects covered.
The book presents the combined research experiences of its 40 authors gathered during a long search in gleaning new knowledge from data. The last page of each chapter has a brief biographical statement of its contributors, who are world-renowned experts.
Audience
The intended audience for this book includes graduate students studying data mining who have some background in mathematical logic and discrete optimization, as well as researchers and practitioners in the same area.
"This book will give the reader a perspective into the core theory and practice of data mining and knowledge discovery (DM & KD). Its chapters combine many theoretical foundations for various DM & KD methods, and they present an array of examples - many of which are drawn from real-life applications. Most of the theoretical developments discussed are accompanied by an extensive empirical analysis, which should give the reader both a deep theoretical and practical insight into the subjects covered." "The intended audience for this book includes graduate students studying data mining who have some background in mathematical logic and discrete optimization as well as researchers and practitioners in the same area."--Jacket This book outlines the core theory and practice of data mining and knowledge discovery (DM & KD) examining theoretical foundations for various methods, and presenting an array of examples, many drawn from real-life applications. Most theoretical developments are accompanied by extensive empirical analysis, offering a deep insight into both theoretical and practical aspects of the subject. The book presents the combined research experiences of 40 expert contributors of world renown.
