Machine Learning Foundations : Supervised, Unsupervised, and Advanced Learning

This book provides conceptual understanding of machine learning algorithms though supervised, unsupervised, and advanced learning techniques. The book consists of four parts: foundation, supervised learning, unsupervised learning, and advanced learning. The first part provides the fundamental materials, background, and simple machine learning algorithms, as the preparation for studying machine learning algorithms. The second and the third parts provide understanding of the supervised learning algorithms and the unsupervised learning algorithms as the core parts. The last part provides advanced machine learning algorithms: ensemble learning, semi-supervised learning, temporal learning, and reinforced learning. * Provides comprehensive coverage of both learning algorithms: supervised and unsupervised learning; * Outlines the computation paradigm for solving classification, regression, and clustering; * Features essential techniques for building the a new generation of machine learning. Preface Part I: Foundation Part II: Supervised Learning Part III: Unsupervised Learning Part IV: Advanced Topics Contents Part I Foundation 1 Introduction 1.1 Definition of Machine Learning 1.2 Application Areas 1.2.1 Classification 1.2.2 Regression 1.2.3 Clustering 1.2.4 Hybrid Tasks 1.3 Machine Learning Types 1.3.1 Supervised Learning 1.3.2 Unsupervised Learning 1.3.3 Semi-supervised Learning 1.3.4 Reinforcement Learning 1.4 Related Areas 1.4.1 Artificial Intelligence 1.4.2 Neural Networks 1.4.3 Data Mining 1.4.4 Soft Computing 1.5 Summary and Further Discussions References 2 Numerical Vectors 2.1 Introduction 2.2 Operations on Numerical Vectors 2.2.1 Definition 2.2.2 Basic Operations 2.2.3 Inner Product 2.2.4 Linear Independence 2.3 Operations on Matrices 2.3.1 Definition 2.3.2 Basic Operations 2.3.3 Multiplication 2.3.4 Inverse Matrix 2.4 Vector and Matrix 2.4.1 Determinant 2.4.2 Eigen Value and Vector 2.4.3 Singular Value Decomposition 2.4.4 Principal Component Analysis 2.5 Summary and Further Discussions 3 Data Encoding 3.1 Introduction 3.2 Relational Data 3.2.1 Basic Concepts 3.2.2 Relational Database 3.2.3 Encoding Process 3.2.4 Encoding Issues 3.3 Textual Data 3.3.1 Text Indexing 3.3.2 Text Encoding 3.3.3 Dimension Reduction 3.3.4 Encoding Issues 3.4 Image Data 3.4.1 Image File Formats 3.4.2 Image Matrix 3.4.3 Encoding Process 3.4.4 Encoding Issues 3.5 Summary and Further Discussions References 4 Simple Machine Learning Algorithms 4.1 Introduction 4.2 Classification 4.2.1 Binary Classification 4.2.2 Multiple Classification 4.2.3 Regression 4.2.4 Problem Decomposition 4.3 Simple Classifiers 4.3.1 Threshold Rule 4.3.2 Rectangle 4.3.3 Hyperplane 4.3.4 Matching Algorithm 4.4 Linear Classifiers 4.4.1 Linear Separability 4.4.2 Hyperplane Equation 4.4.3 Linear Classification 4.4.4 Perceptron 4.5 Summary and Further Discussions References Part II Supervised Learning 5 Instance Based Learning 5.1 Introduction 5.2 Primitive Instance Based Learning 5.2.1 Look-Up Example 5.2.2 Rule Based Approach 5.2.3 Example Similarity 5.2.4 One Nearest Neighbor 5.3 Classification Process 5.3.1 Notations 5.3.2 Nearest Neighbors 5.3.3 Voting 5.3.4 Attribute Discriminations 5.4 Variants 5.4.1 Dynamic Nearest Neighbor 5.4.2 Concentric Nearest Neighbor 5.4.3 Hierarchical Nearest Neighbor 5.4.4 Hub Examples 5.5 Summary and Further Discussions References 6 Probabilistic Learning 6.1 Introduction 6.2 Bayes Classifier 6.2.1 Probabilities 6.2.2 Bayes Rule 6.2.3 Gaussian Distribution 6.2.4 Classification 6.3 Naive Bayes 6.3.1 Classification 6.3.2 Learning 6.3.3 Variants 6.3.4 Application to Text Classification 6.4 Bayesian Learning 6.4.1 Bayesian Networks 6.4.2 Causal Relation 6.4.3 Learning Process 6.4.4 Comparisons 6.5 Summary and Further Discussions References 7 Decision Tree 7.1 Introduction 7.2 Classification Process 7.2.1 Basic Structure 7.2.2 Toy Examples 7.2.3 Text Classification 7.2.4 Rule Extraction 7.3 Learning Process 7.3.1 Preprocessing 7.3.2 Root Node 7.3.3 Interior Nodes 7.3.4 Pruning 7.4 Variants 7.4.1 Regression Version 7.4.2 Decision List 7.4.3 Random Forest 7.4.4 Decision Graph 7.5 Summary and Further Discussions Reference 8 Support Vector Machine 8.1 Introduction 8.2 Classification Process 8.2.1 Linear Classifier 8.2.2 Kernel Functions 8.2.3 Lagrange Multipliers 8.2.4 Generalization 8.3 Learning Process 8.3.1 Primal Problem 8.3.2 Dual Problem 8.3.3 SMO Algorithm 8.3.4 Other Optimization Schemes 8.4 Variants 8.4.1 Fuzzy SVM 8.4.2 Pairwise SVM 8.4.3 LMS SVM 8.4.4 Sparse SVM 8.5 Summary and Further Discussions References Part III Unsupervised Learning 9 Simple Clustering Algorithms 9.1 Introduction 9.2 AHC Algorithm 9.2.1 Cluster Similarity 9.2.2 Initial Version 9.2.3 Fuzzy Clustering 9.2.4 Variants 9.3 Divisive Clustering Algorithm 9.3.1 Binary Clustering 9.3.2 Evolutionary Binary Clustering 9.3.3 Standard Version 9.3.4 Variants 9.4 Online Linear Clustering Algorithm 9.4.1 Representative Selection Scheme 9.4.2 Initial Version 9.4.3 Fuzzy Clustering 9.4.4 Variants 9.5 Summary and Further Discussions References 10 K Means Algorithm 10.1 Introduction 10.2 Supervised and Unsupervised Learning 10.2.1 Learning Paradigm Transition 10.2.2 Unsupervised KNN 10.2.3 Semi-supervised KNN 10.2.4 Dynamic Data Organization 10.3 Clustering Process 10.3.1 Initialization 10.3.2 Hard Clustering 10.3.3 Fuzzy Clustering 10.3.4 Hierarchical Clustering 10.4 Variants 10.4.1 K Medoid Algorithm 10.4.2 Dynamic K Means Algorithm 10.4.3 Semi-supervised Version 10.4.4 Constraint Clustering 10.5 Summary and Further Discussions References 11 EM Algorithm 11.1 Introduction 11.2 Cluster Distributions 11.2.1 Uniform Distribution 11.2.2 Gaussian Distribution 11.2.3 Poisson Distribution 11.2.4 Fuzzy Distributions 11.3 Clustering Process 11.3.1 Initialization 11.3.2 E-Step 11.3.3 M-Step 11.3.4 Issues 11.4 Semi-Supervised Learning: Text Classification 11.4.1 Semi-Supervised Learning 11.4.2 Initialization 11.4.3 Likelihood Estimation 11.4.4 Parameter Estimation 11.5 Summary and Further Discussions References 12 Advanced Clustering 12.1 Introduction 12.2 Cluster Index 12.2.1 Computation Process 12.2.2 Hard Clustering Evaluation 12.2.3 Fuzzy Clustering Evaluation 12.2.4 Hierarchical Clustering Evaluation 12.3 Parameter Tuning 12.3.1 Clustering Index to Unlabeled Items 12.3.2 Simple Clustering Algorithms 12.3.3 K Means Algorithm 12.3.4 Evolutionary Clustering 12.4 Clustering Governance 12.4.1 Cluster Naming 12.4.2 Cluster Maintenance 12.4.3 Multiple Viewed Clustering 12.4.4 Clustering Results Integration 12.5 Summary and Further Discussions References Part IV Advanced Topics 13 Ensemble Learning 13.1 Introduction 13.2 Combination Schemes 13.2.1 Voting 13.2.2 Expert Gates 13.2.3 Cascading 13.2.4 Cellular Model 13.3 Meta-learning 13.3.1 Voting 13.3.2 Expert Gates 13.3.3 Cascading 13.3.4 Cellular Model 13.4 Partition 13.4.1 Training Set Partition 13.4.2 Attribute Set Partition 13.4.3 Architecture Partition 13.4.4 Parallel and Distributed Learning 13.5 Summary and Further Discussions References 14 Semi-supervised Learning 14.1 Introduction 14.2 Kohonen Networks 14.2.1 Initial Version 14.2.2 Learning Vector Quantization 14.2.3 Semi-supervised Version 14.2.4 Kohonen Networks vs. K Means Algorithm 14.3 Combined Learning Algorithms 14.3.1 Combination Paradigms 14.3.2 Simple Learning Algorithms 14.3.3 K Means Algorithm + KNN Algorithm 14.3.4 EM Algorithm + Naive Bayes 14.4 Advanced Supervised Learning 14.4.1 Resampling 14.4.2 Virtual Training Example 14.4.3 Co-Learning 14.4.4 Incremental Learning 14.5 Summary and Further Discussions References 15 Temporal Learning 15.1 Introduction 15.2 Discrete Markov Model 15.2.1 State Diagram 15.2.2 State Transition Probability 15.2.3 State Path Probability 15.2.4 Application to Time Series Prediction 15.3 Hidden Markov Model 15.3.1 Initial Parameters 15.3.2 Observation Sequence Probability 15.3.3 State Sequence Estimation 15.3.4 HMM Learning 15.4 Text Topic Analysis 15.4.1 Task Specification 15.4.2 Sampling 15.4.3 Learning 15.4.4 Topic Sequence 15.5 Summary and Further Discussions References 16 Reinforcement Learning 16.1 Introduction 16.2 Simple Reinforcement Learning 16.2.1 Single Example 16.2.2 Classification 16.2.3 Regression 16.2.4 Autonomous Moving 16.3 Q Learning 16.3.1 Q Table 16.3.2 Finite State 16.3.3 Infinite State 16.3.4 Stochastic Reward 16.4 Advanced Reinforcement Learning 16.4.1 Ensemble Reinforcement Learning 16.4.2 Reinforcement + Supervised 16.4.3 Reinforcement + Unsupervised 16.4.4 Environment Prediction 16.5 Summary and Further Discussions Index