C4.5 (eBook)

Programs for Machine Learning

J. Ross Quinlan (Autor)

eBook Download: PDF

2014 | 1. Auflage
312 Seiten
Elsevier Science (Verlag)
978-0-08-050058-4 (ISBN)

Infos

Classifier systems play a major role in machine learning and knowledge-based systems, and Ross Quinlan's work on ID3 and C4.5 is widely acknowledged to have made some of the most significant contributions to their development. This book is a complete guide to the C4.5 system as implemented in C for the UNIX environment. It contains a comprehensive guide to the system's use , the source code (about 8,800 lines), and implementation notes. C4.5 starts with large sets of cases belonging to known classes. The cases, described by any mixture of nominal and numeric properties, are scrutinized for patterns that allow the classes to be reliably discriminated. These patterns are then expressed as models, in the form of decision trees or sets of if-then rules, that can be used to classify new cases, with emphasis on making the models understandable as well as accurate. The system has been applied successfully to tasks involving tens of thousands of cases described by hundreds of properties. The book starts from simple core learning methods and shows how they can be elaborated and extended to deal with typical problems such as missing data and over hitting. Advantages and disadvantages of the C4.5 approach are discussed and illustrated with several case studies. This book should be of interest to developers of classification-based intelligent systems and to students in machine learning and expert systems courses.

J. Ross Quinlan, University of New South Wales

Inhalt

Front Cover 1
C4.5: Programs for Machine Learning 4
Copyright Page 5
Table of Contents 6
Preface 8
Obtaining the C4.5 Code 10
CHAPTER 1. Introduction 12
1.1 Example: Labor negotiation settlements 14
1.2 Other kinds of classification models 23
1.3 What lies ahead 27
CHAPTER 2. Constructing Decision Trees 28
2.1 Divide and conquer 28
2.2 Evaluating tests 31
2.3 Possible tests considered 35
2.4 Tests on continuous attributes 36
CHAPTER 3. Unknown Attribute Values 38
3.1 Adapting the previous algorithms 39
3.2 Play/Don't Play example again 41
3.3 Recapitulation 43
CHAPTER 4. Pruning Decision Trees 46
4.1 When to simplify? 47
4.2 Error-based pruning 48
4.3 Example: Democrats and Republicans 52
4.4 Estimating error rates for trees 53
CHAPTER 5. From Trees to Rules 56
5.1 Generalizing single rules 58
5.2 Class rulesets 61
5.3 Ranking classes and choosing a default 65
5.4 Summary 66
CHAPTER 6. Windowing 68
6.1 Example: Hypothyroid conditions revisited 69
6.2 Why retain windowing? 69
6.3 Example: The multiplexor 71
CHAPTER 7. Grouping Attribute Values 74
7.1 Finding value groups by merging 75
7.2 Example: Soybean diseases 76
7.3 When to form groups? 77
7.4 Example: The Monk's problems 78
7.5 Uneasy reflections 80
CHAPTER 8. Interacting with Classification Models 82
8.1 Decision tree models 82
8.2 Production rule models 89
8.3 Caveat 91
CHAPTER 9. Guide to Using the System 92
9.1 Files 92
9.2 Running the programs 95
9.3 Conducting experiments 100
9.4 Using options: A credit approval example 102
CHAPTER 10. Limitations 106
10.1 Geometric interpretation 106
10.2 Nonrectangular regions 107
10.3 Poorly delineated regions 109
10.4 Fragmented regions 111
10.5 A more cheerful note 113
CHAPTER 11. Desirable Additions 114
11.1 Continuous classes 114
11.2 Ordered discrete attributes 115
11.3 Structured attributes 115
11.4 Structured induction 116
11.5 Incremental induction 117
11.6 Prospectus 118
Appendix: Program Listings 120
Brief descriptions of the contents of the files 121
Notes on some important data structures 123
File Makefile 126
Alphabetic index of routines 299
References and Bibliography 302
Author Index 308
Subject Index 310

Produkt-Details

• Erscheint lt. Verlag: 28.6.2014
• Sprache: englisch
• Themenwelt: Informatik ► Theorie / Studium ► Künstliche Intelligenz / Robotik
• ISBN-10: 0-08-050058-7 / 0080500587
• ISBN-13: 978-0-08-050058-4 / 9780080500584