Big data and differential privacy : analysis strategies for railway track engineering

Content: Cover
 Title Page
 Copyright
 Contents
 Preface
 Acknowledgments
 Chapter 1 Introduction
 1.1 General
 1.2 Track Components
 1.3 Characteristics of Railway Track Data
 1.4 Railway Track Engineering Problems
 1.5 Wheel-Rail Interface Data
 1.5.1 Switches and Crossings
 1.6 Geometry Data
 1.7 Track Geometry Degradation Models
 1.7.1 Deterministic Models
 1.7.1.1 Linear Models
 1.7.1.2 Nonlinear Models
 1.7.2 Stochastic Models
 1.7.3 Discussion
 1.8 Rail Defect Data
 1.9 Inspection and Detection Systems
 1.10 Rail Grinding
 1.11 Traditional Data Analysis Techniques
 1.11.1 Emerging Data Analysis. 1.12 RemarksReferences
 Chapter 2 Data Analysis --
Basic Overview
 2.1 Introduction
 2.2 Exploratory Data Analysis (EDA)
 2.3 Symbolic Data Analysis
 2.3.1 Building Symbolic Data
 2.3.2 Advantages of Symbolic Data
 2.4 Imputation
 2.5 Bayesian Methods and Big Data Analysis
 2.6 Remarks
 References
 Chapter 3 Machine Learning: A Basic Overview
 3.1 Introduction
 3.2 Supervised Learning
 3.3 Unsupervised Learning
 3.4 Semi-Supervised Learning
 3.5 Reinforcement Learning
 3.6 Data Integration
 3.7 Data Science Ontology
 3.7.1 Kernels
 3.7.1.1 General
 3.7.1.2 Learning Process. 3.7.2 Basic Operations with Kernels3.7.3 Different Kernel Types
 3.7.4 Intuitive Example
 3.7.5 Kernel Methods
 3.7.5.1 Support Vector Machines
 3.8 Imbalanced Classification
 3.9 Model Validation
 3.9.1 Receiver Operating Characteristic (ROC) Curves
 3.9.1.1 ROC Curves
 3.10 Ensemble Methods
 3.10.1 General
 3.10.2 Bagging
 3.10.3 Boosting
 3.11 Big P and Small N (P k N)
 3.11.1 Bias and Variances
 3.11.2 Multivariate Adaptive Regression Splines (MARS)
 3.12 Deep Learning
 3.12.1 General
 3.12.2 Deep Belief Networks
 3.12.2.1 Restricted Boltzmann Machines (RBM). 3.12.2.2 Deep Belief Nets (DBN)3.12.3 Convolutional Neural Networks (CNN)
 3.12.4 Granular Computing (Rough Set Theory)
 3.12.5 Clustering
 3.12.5.1 Measures of Similarity or Dissimilarity
 3.12.5.2 Hierarchical Methods
 3.12.5.3 Non-Hierarchical Clustering
 3.12.5.4 k-Means Algorithm
 3.12.5.5 Expectation-Maximization (EM) Algorithms
 3.13 Data Stream Processing
 3.13.1 Methods and Analysis
 3.13.2 LogLog Counting
 3.13.3 Count-Min Sketch
 3.13.3.1 Online Support Regression
 3.14 Remarks
 References
 Chapter 4 Basic Foundations of Big Data
 4.1 Introduction
 4.2 Query. 4.3 Taxonomy of Big Data Analytics in Railway Track Engineering4.4 Data Engineering
 4.5 Remarks
 References
 Chapter 5 Hilbert-Huang Transform, Profile, Signal, and Image Analysis
 5.1 Hilbert-Huang Transform
 5.1.1 Traditional Empirical Mode Decomposition
 5.1.1.1 Side Effect (Boundary Effect)
 5.1.1.2 Example
 5.1.1.3 Stopping Criterion
 5.1.2 Ensemble Empirical Mode Decomposition (EEMD)
 5.1.2.1 Post-Processing EEMD
 5.1.3 Complex Empirical Mode Decomposition (CEMD)
 5.1.4 Spectral Analysis
 5.1.5 Bidimensional Empirical Mode Decomposition (BEMD)
 5.1.5.1 Example.