دانلود کتاب Data Science: Theory and Applications (Volume 44) (Handbook of Statistics, Volume 44)

فهرست مطالب :

Front Cover
Data Science: Theory and Applications
Copyright
Contents
Contributors
Preface
Section I: Animal models and ecological large data methods
Chapter 1: Statistical outline of animal home ranges: An application of set estimation
1. Introduction to home range estimation
1.1. Problem statement
1.2. Connection to the set estimation problem
1.2.1. Support estimation
1.2.2. Level set estimation
1.3. Characteristics of animal location data
1.4. A real data set: Mongolian wolves
2. Statistical techniques for home range estimation
2.1. Assuming location data are independent
2.1.1. Global methods
2.1.1.1. Minimum convex polygon (MCP) or convex hull
2.1.1.2. α-Convex hull
2.1.1.3. Kernel density estimation (KDE)
2.1.1.4. Harmonic mean home range
2.1.2. Localized methods
2.1.2.1. Local convex hull (LoCoH) or k-nearest neighbor convex hulls (k-NNCH)
2.1.2.2. Characteristic hull polygon (CHP)
2.1.2.3. Single-linkage cluster
2.2. Incorporating time dependency
2.2.1. Global methods
2.2.1.1. Time geography density estimation (TGDE)
2.2.1.2. Kernel density estimation
2.2.2. Localized methods
2.2.2.1. T-LoCoH
3. Developing trends, open problems, and future work
3.1. Incorporating time dependence into the home range estimator
3.2. Taking advantage of explanatory variables
3.3. Selecting the ``optimal´´ home range
Acknowledgments
References
Chapter 2: Modeling extreme climatic events using the generalized extreme value (GEV) distribution
1. Introduction
1.1. Previous models of extreme events in ecology
2. The GEV value distribution
2.1. A brief history of the GEV
2.2. Definitions
2.3. Parameterizing the GEV
2.4. Return level vs return period
2.5. Evaluating model fit
2.6. Applying climate change to the GEV
3. Case study: Hurricanes and Ardisia escallonioides
3.1. Traditional hurricane modeling
3.2. The data: HURDAT2
3.3. Study region and storm selection criteria
3.4. Fitting the GEV model to our case study
3.5. The demographic model
3.6. Constructing the environmental transition matrix (P)
3.7. Incorporating climate change in the environmental transition matrix (P)
3.8. Population simulations
3.9. Sensitivity analysis
3.10. Effects of changing hurricane frequency, intensity, damage levels, and/or canopy recovery rates
4. Challenges
4.1. Defining extreme events
4.2. Measuring the impacts of extreme events
5. Open questions and future applications
6. Summary
7. Code availability
Acknowledgments
References
Section II: Engineering sciences data
Chapter 3: Blockchain technology: Theory and practice
1. Introduction
1.1. Bitcoin network
1.2. Blockchain
1.3. How is blockchain different from present database systems
1.3.1. Distributed consensus algorithm
1.3.2. Organization of the chapter
2. Cryptographic foundations
2.1. Secure hash function
2.1.1. A note on computational hardness
2.2. Open issue in theoretical computer science
2.3. Encryption methods
2.3.1. Public key encryption system
2.3.2. Authenticity of public key and private key pairs
2.3.3. Symmetric key encryption system
2.3.4. Comparison
2.3.5. Usage in blockchain network
2.4. Digital signature mechanism
2.5. Pseudo random number generators
2.5.1. Appending random numbers to messages
2.6. Zero-knowledge protocols
2.6.1. Formal notion
2.6.2. Interactive proof (example 1)
2.6.3. Interactive proof (example 2)
2.6.4. Noninteractive proof (example)
3. Blockchain platforms
3.1. Commonalities of public and private blockchains
3.2. Consistency of Ledger across the network
3.2.1. Consensus algorithm
4. Public blockchains
4.1. Bitcoin network
4.1.1. Proof-of-Work (PoW) algorithm
4.1.2. Miners
4.1.3. Cost of successful mining
4.2. Ethereum network
4.2.1. Proof-of-work
4.2.2. Proof-of-stake
4.2.3. Smart contracts
5. Private blockchains
5.1. Hyperledger fabric
5.1.1. Transaction cycle
5.1.2. Endorsement policy
5.1.3. Modularity
5.2. Applications
5.2.1. Banking sector
5.2.2. Supply chains
5.3. Discussion and conclusions
References
Further reading
Chapter 4: Application of data handling techniques to predict pavement performance
1. Introduction
2. Importance of data in pavement designs
3. Challenges with pavement data handling
4. Data handling techniques in pavement analysis
5. Data handling and automation framework for pavement design
5.1. Data generation
5.2. Data appending
5.3. Automation
5.4. Data extraction
6. Efficiency of the automation-based data handling framework
7. Regression analysis using the generated data
8. Validation of proposed damage models
9. Summary and recommendations
References
Section III: Statistical estimation designs: Fractional fields, biostatistics and non-parametrics
Chapter 5: On the usefulness of lattice approximations for fractional Gaussian fields
1. Introduction
2. Fractional Gaussian fields and their approximations
2.1. Fractional Gaussian fields
2.2. Lattice approximations
3. Model based geostatistics
3.1. Maximum likelihood estimation
3.1.1. MLE for fractional Gaussian fields
3.1.2. MLE with lattice approximations
4. Simulation studies
4.1. An experiment with power-law variogram
4.2. Large scale computation with lattice approximations
5. Indian ocean surface temperature from Argo floats
6. Concluding remarks
Acknowledgments
Appendix
MATLAB codes for Section 3.1.1
MATLAB codes for Section 3.1.2
References
Chapter 6: Estimating individual-level average treatment effects: Challenges, modeling approaches, and practical applications
1. Introduction
2. Introduction to statistical learning
2.1. Preliminaries
2.2. Bias-variance decomposition
2.3. Examples of the bias-variance tradeoff
2.4. Regularization
2.5. Evaluating a model\'s prediction performance
3. Causal inference
3.1. Notation and counterfactual theory
3.2. Theoretical MSE criteria for causal inference
4. Challenges in learning the CATE function
4.1. The naive optimization strategy fails
4.1.1. Generalized random forests
4.1.2. Example: Difference of independently fit μ1(x) and μ0(x) does not achieve performance of single fit τ(x)
4.2. Selection bias
4.3. Regularization biases
5. Survey of CATE modeling approaches
5.1. Meta-learners
5.2. Transformed outcome methods
5.3. Shared basis methods
5.4. Neural networks
5.5. Bayesian nonparametric methods
6. CATE model selection and assessment
6.1. Survey of evaluation methods for estimators of τ(x)
7. Practical application of CATE estimates
7.1. Stratification on the CATEs
7.2. Subgroup identification
8. Summary
References
Chapter 7: Nonparametric data science: Testing hypotheses in large complex data
1. Introduction
2. Nonparametric data science
3. One-sample methods
4. Two-sample methods
5. Multi-sample methods
6. Multivariate methods
7. Ranked-set based methods
8. Changepoint detection
9. Final remarks
References
Section IV: Network models and COVID-19 modeling
Chapter 8: Network models in epidemiology
1. Introduction
2. Network models
2.1. Static network
2.1.1. Basic reproduction number
2.2. Dynamic network
2.2.1. Correlation and dynamics
2.2.2. SIR model in network
2.3. Multi-layered model
3. Syringe exchange on Hepatitis C
3.1. Syringe exchange SIRS model
3.2. Syringe exchange SIRS model adjusted with network
4. Discussion
References
Chapter 9: Modeling and forecasting the spread of COVID-19 pandemic in India and significance of lockdown: A mathematica
1. Background
2. Why mathematical modeling?
2.1. SIR and SEIR model
3. Formulation of SEIR model
3.1. Growth models
3.1.1. Linear growth
3.1.2. Exponential growth (unlimited population growth)
3.1.3. Logistic growth (sigmoidal)
3.2. Significance of lockdown
3.3. Propagation model (based on Newton\'s law of cooling)
3.4. Joinpoint regression model
4. Conclusions
References
Further reading
Chapter 10: Mathematical modeling as a tool for policy decision making: Applications to the COVID-19 pandemic
1. Introduction
1.1. Overview of mathematical modeling
1.2. Modeling to explain or to predict
1.3. What does it mean for models to be ``right´´?
1.4. Aims and purposes of this work
2. Modeling of infectious disease
2.1. The essence of modeling infectious diseases
2.2. History of modeling infectious diseases
2.2.1. Lotka-Volterra equations, SIR models and reproduction number
2.2.2. Stochastic models: Adding stochasticity to the SIR framework, branching processes and individual- or agent-based m ...
2.3. Challenges of modeling infectious diseases
3. Models for COVID-19
3.1. Compartmental modeling and the SEIR framework: Overview and examples of COVID-19 models
3.2. Agent-based models: Overview and examples of COVID-19 models
3.3. Branching process models: Overview and examples of COVID-19 models
4. Applications of modeling of COVID-19: Three case studies
4.1. Case study 1: Application of SEIR-TTI model to the UK COVID-19 epidemic
4.1.1. Overview of SEIR-TTI
4.1.2. SEIR-TTI methodology
4.1.3. Application of SEIR-TTI to estimate R
4.2. Case study 2: Application of Covasim to the UK COVID-19 epidemic
4.2.1. Overview of Covasim
4.2.2. Covasim methodology
4.2.3. Application of Covasim to the UK epidemic
4.3. Case study 3: Application of rule-based modeling
4.3.1. Overview
4.3.2. Rule-based modeling methodology
4.3.3. Application of RBM to LMICs
5. Conclusions
Acknowledgments
References
Index
Back Cover