第1页 / 共348页
第2页 / 共348页
第3页 / 共348页
第4页 / 共348页
第5页 / 共348页
第6页 / 共348页
第7页 / 共348页
第8页 / 共348页
An Introduction to Machine Learning(2nd) 无水印pdf.pdf
Cover
Copyright
Contents
Introduction
1 A Simple Machine-Learning Task
1.1 Training Sets and Classifiers
What Have You Learned?
1.2 Minor Digression: Hill-Climbing Search
What Have You Learned?
1.3 Hill Climbing in Machine Learning
What Have You Learned?
1.4 The Induced Classifier's Performance
What Have You Learned?
1.5 Some Difficulties with Available Data
What Have You Learned?
1.6 Summary and Historical Remarks
1.7 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
2 Probabilities: Bayesian Classifiers
2.1 The Single-Attribute Case
What Have You Learned?
2.2 Vectors of Discrete Attributes
What Have You Learned?
2.3 Probabilities of Rare Events: Exploiting the Expert's Intuition
What Have You Learned?
2.4 How to Handle Continuous Attributes
What Have You Learned?
2.5 Gaussian ``Bell'' Function: A Standard pdf
What Have You Learned?
2.6 Approximating PDFs with Sets of Gaussians
What Have You Learned?
2.7 Summary and Historical Remarks
2.8 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
3 Similarities: Nearest-Neighbor Classifiers
3.1 The k-Nearest-Neighbor Rule
What Have You Learned?
3.2 Measuring Similarity
What Have You Learned?
3.3 Irrelevant Attributes and Scaling Problems
What Have You Learned?
3.4 Performance Considerations
What Have You Learned?
3.5 Weighted Nearest Neighbors
What Have You Learned?
3.6 Removing Dangerous Examples
What Have You Learned?
3.7 Removing Redundant Examples
What Have You Learned?
3.8 Summary and Historical Remarks
3.9 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
4 Inter-Class Boundaries: Linear and Polynomial Classifiers
4.1 The Essence
What Have You Learned?
4.2 The Additive Rule: Perceptron Learning
What Have You Learned?
4.3 The Multiplicative Rule: WINNOW
What Have You Learned?
4.4 Domains with More Than Two Classes
What Have You Learned?
4.5 Polynomial Classifiers
What Have You Learned?
4.6 Specific Aspects of Polynomial Classifiers
What Have You Learned?
4.7 Numerical Domains and Support Vector Machines
What Have You Learned?
4.8 Summary and Historical Remarks
4.9 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
5 Artificial Neural Networks
5.1 Multilayer Perceptrons as Classifiers
What Have You Learned?
5.2 Neural Network's Error
What Have You Learned?
5.3 Backpropagation of Error
What Have You Learned?
5.4 Special Aspects of Multilayer Perceptrons
What Have You Learned?
5.5 Architectural Issues
What Have You Learned?
5.6 Radial-Basis Function Networks
What Have You Learned?
5.7 Summary and Historical Remarks
5.8 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
6 Decision Trees
6.1 Decision Trees as Classifiers
What Have You Learned?
6.2 Induction of Decision Trees
What Have You Learned?
6.3 How Much Information Does an Attribute Convey?
What Have You Learned?
6.4 Binary Split of a Numeric Attribute
What Have You Learned?
6.5 Pruning
What Have You Learned?
6.6 Converting the Decision Tree into Rules
What Have You Learned?
6.7 Summary and Historical Remarks
6.8 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
7 Computational Learning Theory
7.1 PAC Learning
What Have You Learned?
7.2 Examples of PAC Learnability
What Have You Learned?
7.3 Some Practical and Theoretical Consequences
What Have You Learned?
7.4 VC-Dimension and Learnability
What Have You Learned?
7.5 Summary and Historical Remarks
7.6 Exercises and Thought Experiments
Exercises
Give It Some Thought
8 A Few Instructive Applications
8.1 Character Recognition
What Have You Learned?
8.2 Oil-Spill Recognition
What Have You Learned?
8.3 Sleep Classification
What Have You Learned?
8.4 Brain–Computer Interface
What Have You Learned?
8.5 Medical Diagnosis
What Have You Learned?
8.6 Text Classification
What Have You Learned?
8.7 Summary and Historical Remarks
8.8 Exercises and Thought Experiments
Give It Some Thought
Computer Assignments
9 Induction of Voting Assemblies
9.1 Bagging
What Have You Learned?
9.2 Schapire's Boosting
What Have You Learned?
9.3 Adaboost: Practical Version of Boosting
What Have You Learned?
9.4 Variations on the Boosting Theme
What Have You Learned?
9.5 Cost-Saving Benefits of the Approach
What Have You Learned?
9.6 Summary and Historical Remarks
9.7 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
10 Some Practical Aspects to Know About
10.1 A Learner's Bias
What Have You Learned?
10.2 Imbalanced Training Sets
What Have You Learned?
10.3 Context-Dependent Domains
What Have You Learned?
10.4 Unknown Attribute Values
What Have You Learned?
10.5 Attribute Selection
What Have You Learned?
10.6 Miscellaneous
What Have You Learned?
10.7 Summary and Historical Remarks
10.8 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
11 Performance Evaluation
11.1 Basic Performance Criteria
What Have You Learned?
11.2 Precision and Recall
What Have You Learned?
11.3 Other Ways to Measure Performance
What Have You Learned?
11.4 Learning Curves and Computational Costs
What Have You Learned?
11.5 Methodologies of Experimental Evaluation
What Have You Learned?
11.6 Summary and Historical Remarks
11.7 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
12 Statistical Significance
12.1 Sampling a Population
What Have You Learned?
12.2 Benefiting from the Normal Distribution
What Have You Learned?
12.3 Confidence Intervals
What Have You Learned?
12.4 Statistical Evaluation of a Classifier
What Have You Learned?
12.5 Another Kind of Statistical Evaluation
What Have You Learned?
12.6 Comparing Machine-Learning Techniques
What Have You Learned?
12.7 Summary and Historical Remarks
12.8 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
13 Induction in Multi-Label Domains
13.1 Classical Machine Learning in Multi-Label Domains
What Have You Learned?
13.2 Treating Each Class Separately: Binary Relevance
What Have You Learned?
13.3 Classifier Chains
What Have You Learned?
13.4 Another Possibility: Stacking
What Have You Learned?
13.5 A Note on Hierarchically Ordered Classes
What Have You Learned?
13.6 Aggregating the Classes
What Have You Learned?
13.7 Criteria for Performance Evaluation
What Have You Learned?
13.8 Summary and Historical Remarks
13.9 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
14 Unsupervised Learning
14.1 Cluster Analysis
What Have You Learned?
14.2 A Simple Algorithm: k-Means
What Have You Learned?
14.3 More Advanced Versions of k-Means
What Have You Learned?
14.4 Hierarchical Aggregation
What Have You Learned?
14.5 Self-Organizing Feature Maps: Introduction
What Have You Learned?
14.6 Some Important Details
What Have You Learned?
14.7 Why Feature Maps?
What Have You Learned?
14.8 Summary and Historical Remarks
14.9 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
15 Classifiers in the Form of Rulesets
15.1 A Class Described By Rules
What Have You Learned?
15.2 Inducing Rulesets by Sequential Covering
What Have You Learned?
15.3 Predicates and Recursion
What Have You Learned?
15.4 More Advanced Search Operators
What Have You Learned?
15.5 Summary and Historical Remarks
15.6 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
16 The Genetic Algorithm
16.1 The Baseline Genetic Algorithm
What Have You Learned?
16.2 Implementing the Individual Modules
What Have You Learned?
16.3 Why It Works
What Have You Learned?
16.4 The Danger of Premature Degeneration
What Have You Learned?
16.5 Other Genetic Operators
What Have You Learned?
16.6 Some Advanced Versions
What Have You Learned?
16.7 Selections in k-NN Classifiers
What Have You Learned?
16.8 Summary and Historical Remarks
16.9 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
17 Reinforcement Learning
17.1 How to Choose the Most Rewarding Action
What Have You Learned?
17.2 States and Actions in a Game
What Have You Learned?
17.3 The SARSA Approach
What Have You Learned?
17.4 Summary and Historical Remarks
17.5 Solidify Your Knowledge
Exercises
Give It Some Thought
Computer Assignments
Bibliography
Index
Miroslav Kubat
An Introduction
to Machine
Learning
Second Edition
An Introduction to Machine Learning
Miroslav Kubat
An Introduction to Machine
Learning
Second Edition
123
Miroslav Kubat
Department of Electrical and Computer Engineering
University of Miami
Coral Gables, FL, USA
ISBN 978-3-319-63912-3
DOI 10.1007/978-3-319-63913-0
ISBN 978-3-319-63913-0 (eBook)
Library of Congress Control Number: 2017949183
© Springer International Publishing AG 2015, 2017
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
Printed on acid-free paper
This Springer imprint is published by Springer Nature
The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To my wife, Verunka.
Contents
1
2
3
A Simple Machine-Learning Task ........................................
1.1
Training Sets and Classifiers .........................................
1.2 Minor Digression: Hill-Climbing Search ...........................
Hill Climbing in Machine Learning .................................
1.3
The Induced Classifier’s Performance...............................
1.4
Some Difficulties with Available Data ..............................
1.5
1.6
Summary and Historical Remarks ...................................
Solidify Your Knowledge ............................................
1.7
Probabilities: Bayesian Classifiers ........................................
The Single-Attribute Case ...........................................
2.1
2.2
Vectors of Discrete Attributes .......................................
Probabilities of Rare Events: Exploiting the Expert’s Intuition ....
2.3
How to Handle Continuous Attributes ..............................
2.4
Gaussian “Bell” Function: A Standard pdf .........................
2.5
2.6
Approximating PDFs with Sets of Gaussians .......................
Summary and Historical Remarks ...................................
2.7
2.8
Solidify Your Knowledge ............................................
Similarities: Nearest-Neighbor Classifiers ...............................
3.1
The k-Nearest-Neighbor Rule .......................................
3.2 Measuring Similarity .................................................
Irrelevant Attributes and Scaling Problems .........................
3.3
3.4
Performance Considerations .........................................
3.5 Weighted Nearest Neighbors ........................................
Removing Dangerous Examples.....................................
3.6
Removing Redundant Examples.....................................
3.7
Summary and Historical Remarks ...................................
3.8
3.9
Solidify Your Knowledge ............................................
1
1
5
8
11
13
15
16
19
19
22
26
30
33
34
36
40
43
43
46
49
52
55
57
59
61
62
vii
viii
4
5
6
7
8
Contents
Inter-Class Boundaries: Linear and Polynomial Classifiers ...........
The Essence ..........................................................
4.1
The Additive Rule: Perceptron Learning ............................
4.2
4.3
The Multiplicative Rule: WINNOW ................................
Domains with More Than Two Classes .............................
4.4
Polynomial Classifiers ...............................................
4.5
Specific Aspects of Polynomial Classifiers .........................
4.6
4.7
Numerical Domains and Support Vector Machines ................
Summary and Historical Remarks ...................................
4.8
4.9
Solidify Your Knowledge ............................................
65
65
69
73
76
79
81
84
86
87
Artificial Neural Networks ................................................
91
91
5.1 Multilayer Perceptrons as Classifiers ................................
95
Neural Network’s Error ..............................................
5.2
Backpropagation of Error ............................................
5.3
97
Special Aspects of Multilayer Perceptrons.......................... 100
5.4
5.5
Architectural Issues .................................................. 104
Radial-Basis Function Networks .................................... 106
5.6
Summary and Historical Remarks ................................... 109
5.7
5.8
Solidify Your Knowledge ............................................ 110
Decision Trees ............................................................... 113
Decision Trees as Classifiers......................................... 113
6.1
Induction of Decision Trees ......................................... 117
6.2
How Much Information Does an Attribute Convey? ............... 119
6.3
6.4
Binary Split of a Numeric Attribute ................................. 122
Pruning................................................................ 126
6.5
Converting the Decision Tree into Rules ............................ 130
6.6
Summary and Historical Remarks ................................... 132
6.7
6.8
Solidify Your Knowledge ............................................ 133
Computational Learning Theory ......................................... 137
PAC Learning......................................................... 137
7.1
Examples of PAC Learnability ...................................... 141
7.2
7.3
Some Practical and Theoretical Consequences ..................... 143
VC-Dimension and Learnability..................................... 145
7.4
Summary and Historical Remarks ................................... 148
7.5
7.6
Exercises and Thought Experiments ................................ 149
A Few Instructive Applications ........................................... 151
Character Recognition ............................................... 151
8.1
Oil-Spill Recognition ................................................ 155
8.2
8.3
Sleep Classification .................................................. 158
8.4
Brain–Computer Interface ........................................... 161
8.5 Medical Diagnosis.................................................... 165
Contents
ix
8.6
8.7
8.8
Text Classification .................................................... 167
Summary and Historical Remarks ................................... 169
Exercises and Thought Experiments ................................ 170
9
Induction of Voting Assemblies ........................................... 173
Bagging ............................................................... 173
9.1
Schapire’s Boosting .................................................. 176
9.2
Adaboost: Practical Version of Boosting ............................ 179
9.3
9.4
Variations on the Boosting Theme................................... 183
Cost-Saving Benefits of the Approach .............................. 185
9.5
Summary and Historical Remarks ................................... 187
9.6
9.7
Solidify Your Knowledge ............................................ 188
10 Some Practical Aspects to Know About.................................. 191
10.1 A Learner’s Bias...................................................... 191
10.2
Imbalanced Training Sets ............................................ 194
10.3 Context-Dependent Domains ........................................ 199
10.4 Unknown Attribute Values ........................................... 202
10.5 Attribute Selection ................................................... 204
10.6 Miscellaneous ........................................................ 206
10.7 Summary and Historical Remarks ................................... 208
10.8 Solidify Your Knowledge ............................................ 208
11 Performance Evaluation ................................................... 211
11.1 Basic Performance Criteria .......................................... 211
11.2 Precision and Recall.................................................. 214
11.3 Other Ways to Measure Performance ............................... 219
11.4 Learning Curves and Computational Costs ......................... 222
11.5 Methodologies of Experimental Evaluation ........................ 224
11.6 Summary and Historical Remarks ................................... 227
11.7 Solidify Your Knowledge ............................................ 228
12 Statistical Significance ..................................................... 231
12.1 Sampling a Population ............................................... 231
12.2 Benefiting from the Normal Distribution ........................... 235
12.3 Confidence Intervals ................................................. 239
12.4 Statistical Evaluation of a Classifier ................................. 241
12.5 Another Kind of Statistical Evaluation .............................. 244
12.6 Comparing Machine-Learning Techniques ......................... 245
12.7 Summary and Historical Remarks ................................... 247
12.8 Solidify Your Knowledge ............................................ 248
13
Induction in Multi-Label Domains ....................................... 251
13.1 Classical Machine Learning in Multi-Label Domains .............. 251
13.2 Treating Each Class Separately: Binary Relevance ................. 254
13.3 Classifier Chains ..................................................... 256
相关推荐
2023年江西萍乡中考道德与法治真题及答案.doc
2012年重庆南川中考生物真题及答案.doc
2013年江西师范大学地理学综合及文艺理论基础考研真题.doc
2020年四川甘孜小升初语文真题及答案I卷.doc
2020年注册岩土工程师专业基础考试真题及答案.doc
2023-2024学年福建省厦门市九年级上学期数学月考试题及答案.doc
2021-2022学年辽宁省沈阳市大东区九年级上学期语文期末试题及答案.doc
2022-2023学年北京东城区初三第一学期物理期末试卷及答案.doc
2018上半年江西教师资格初中地理学科知识与教学能力真题及答案.doc
2012年河北国家公务员申论考试真题及答案-省级.doc
2020-2021学年江苏省扬州市江都区邵樊片九年级上学期数学第一次质量检测试题及答案.doc
2022下半年黑龙江教师资格证中学综合素质真题及答案.doc
