《迁移学习 理论与实践》PDF下载

  • 购买积分:8 如何计算积分?
  • 作  者:邵浩著
  • 出 版 社:上海:上海交通大学出版社
  • 出版年份:2013
  • ISBN:9787313106568
  • 页数:121 页
图书介绍:本书着眼于管理实际中的资源再利用,对数据挖掘领域最前沿的迁移学习(transfer learning)进行了详细阐述,并着重介绍了应用最为广泛的分类学习(classification),将最前沿的研究进行了归纳总结,并通过实际算法分析,提供给读者领域内的最新进展,使读者能够使用迁移学习的工具构建模型并应用到实际问题。该书主要读者对象为具有管理和计算机背景并在数据挖掘领域有初步研究的学者。

Chapter 1 Introduction 1

1.1 Background and Motivation 1

1.2 Contributions 5

1.2.1 Extended MDLP for Transfer Learning 5

1.2.2 Compact Coding for Hyperplane Classifiers in Transfer Learning 6

1.2.3 Transfer Active Learning 7

1.2.4 Gaussian Process for Transfer Learning 8

1.3 Book Overview 9

Chapter 2 Literature Review and Preliminaries for MDLP 10

2.1 Transfer Learning 10

2.2 Active Learning and Transfer Active Learning 13

2.3 Preliminaries for MDLP 14

Chapter 3 Extended MDL Principle for Feature-based Transfer Learning 17

3.1 Introduction 17

3.2 Problem Statement 20

3.3 Preliminaries for Encoding 21

3.3.1 Theoretical Foundation of the EMDLP 22

3.3.2 Adaptation of the EMDLP to Our Problem 25

3.4 Supervised Inductive Transfer Learning Algorithm 30

3.4.1 EMDLP with Incremental Search 30

3.4.2 EMDLP with Hill Climbing 33

3.5 Experiments 36

3.5.1 Experimental Settings 36

3.5.2 Experimental Results on Synthetic Data Sets 40

3.5.3 Experimental Results on Real Data Sets 45

3.6 Summary 52

Chapter 4 Compact Coding for Hyperplane Classifiers in a Heterogeneous Environment 53

4.1 Introduction 53

4.2 Problem Setting 55

4.3 Compact Coding for Hyperplane Classifiers in Heterogeneous Environment 56

4.3.1 Macro Level:Arrange Related Tasks 57

4.3.2 Micro Level Evaluation 61

4.3.3 The Transfer Learning Algorithm 62

4.4 Experiments 63

4.4.1 Experimental Setting 63

4.4.2 Experimental Results 65

4.5 Summary 71

Chapter 5 Adaptive Transfer Learning with Query by Committee 72

5.1 Introduction 72

5.2 Problem Setting and Preliminaries 75

5.3 Probabilistic Framework for ALTL 78

5.4 The ALTL Algorithm and Analysis 81

5.4.1 The Procedure of ALTL 81

5.4.2 Termination Condition and Analysis 83

5.5 Experiments 85

5.5.1 Experimental Setting 85

5.5.2 Results on Synthetic Data Sets 85

5.5.3 Results on Real Data Sets 89

5.6 Summary 93

Chapter 6 Gaussian Process for Transfer Learning through Minimum Encoding 94

6.1 Introduction 94

6.2 Gaussian Process for Classification 96

6.3 The GPTL Algorithm 97

6.3.1 Arrange Related Tasks 97

6.3.2 The Instance Level Similarities 99

6.4 Experiments 100

6.5 Summary 104

Chapter 7 Concluding Comments 106

Appendix A Target Concepts in Chapter 3 110

Bibliography 113