This class-tested textbook is designed for a semester-long graduate, or senior undergraduate course on Computational Health Informatics. The focus of the book is on computational techniques that are widely used in health data analysis and health informatics. It integrates computer science and clinical perspectives. The book prepares computer science students for careers in computational health informatics and medical data analysis. Features Integrates computer science and clinical perspectives. Describes various statistical and artificial intelligence techniques including machine learning techniques such as clustering including temporal data, regression analysis, neural networks, HMM, decision trees, SVM, and data mining widely used in health-data analysis. Describes computational techniques such as multidimensional and multimedia data representation and retrieval, ontology, patient-data deidentification, temporal data analysis, heterogeneous databases, medical image analysis and transmission, biosignal analysis, pervasive healthcare, automated text-analysis, health-vocabulary knowledgebases and medical information-exchange. Includes bioinformatics and pharmacokinetics techniques and their applications to vaccine and drug development. Arvind Bansal is a full professor of Computer Science at Kent State University, Kent, Ohio, USA. He received his PhD (1988) from Case Western Reserve University, Cleveland, Ohio, USA. His research publications, and undergraduate and graduate teaching are in artificial intelligence, multimedia systems and languages, bioinformatics, and computational health informatics. Javed Khan is a full professor of Computer Science at Kent State University, Kent, Ohio, USA. He received his PhD (1995) from University of Hawaii at Manoa, USA. His research publications, and undergraduate and graduate teachings are in artificial intelligence, computer networking protocols, educational networks, medical image processing and communication, perceptual enhancement, and automated knowledge acquisition. He has been a long-term Fulbright area expert. S. Kaisar Alam received his PhD (1996) in Electrical Engineering from University of Rochester, Rochester NY, USA. His research publications and teaching are in medical image analysis and genome analysis. He was a member of the research staff in Biomedical Engineering Laboratories during 1998-2013. He has been a Fullbright scholar and a visiting professor at RUTGERS University, NY, USA. Currently, he runs his company for medical image analysis.
Author(s): Silvelyn Zwanzig; Behrang Mahjani
Publisher: CRC Press
Year: 2020
Language: English
Pages: xiv+212
Cover
Half Title
Title Page
Copyright Page
Contents
Preface
Introduction
1. Random Variable Generation
1.1 Basic Methods
1.1.1 Congruential Generators
1.1.2 The KISS Generator
1.1.3 Beyond Uniform Distributions
1.2 Transformation Methods
1.3 Accept-Reject Methods
1.3.1 Envelope Accept-Reject Methods
1.4 Problems
2. Monte Carlo Methods
2.1 Independent Monte Carlo Methods
2.1.1 Importance Sampling
2.1.2 The Rule of Thumb for Importance Sampling
2.2 Markov Chain Monte Carlo
2.2.1 Metropolis-Hastings Algorithm
2.2.2 Special MCMC Algorithms
2.2.3 Adaptive MCMC
2.2.4 Perfect Simulation
2.2.5 The Gibbs Sampler
2.3 Approximate Bayesian Computation Methods
2.4 Problems
3. Bootstrap
3.1 General Principle
3.1.1 Unified Bootstrap Framework
3.1.2 Bootstrap and Monte Carlo
3.1.3 Conditional and Unconditional Distribution
3.2 Basic Bootstrap
3.2.1 Plug-in Principle
3.2.2 Why is Bootstrap Good?
3.2.3 Example where Bootstrap Fails
3.3 Bootstrap Confidence Sets
3.3.1 The Pivotal Method
3.3.2 Bootstrap Pivotal Methods
3.3.2.1 Percentile Bootstrap Confidence Interval
3.3.2.2 Basic Bootstrap Confidence Interval
3.3.2.3 Studentized Bootstrap Confidence Interval
3.3.3 Transformed Bootstrap Confidence Intervals
3.3.4 Prepivoting Confidence Set
3.3.5 BCa-Confidence Interval
3.4 Bootstrap Hypothesis Tests
3.4.1 Parametric Bootstrap Hypothesis Test
3.4.2 Nonparametric Bootstrap Hypothesis Test
3.4.3 Advanced Bootstrap Hypothesis Tests
3.5 Bootstrap in Regression
3.5.1 Model-Based Bootstrap
3.5.2 Parametric Bootstrap Regression
3.5.3 Casewise Bootstrap in Correlation Model
3.6 Bootstrap for Time Series
3.7 Problems
4. Simulation-Based Methods
4.1 EM Algorithm
4.2 SIMEX
4.3 Variable Selection
4.3.1 F-Backward and F-Forward Procedures
4.3.2 FSR-Forward Procedure
4.3.3 SimSel
4.4 Problems
5. Density Estimation
5.1 Background
5.2 Histogram
5.3 Kernel Density Estimator
5.3.1 Statistical Properties
5.3.2 Bandwidth Selection in Practice
5.4 Nearest Neighbor Estimator
5.5 Orthogonal Series Estimator
5.6 Minimax Convergence Rate
5.7 Problems
6. Nonparametric Regression
6.1 Background
6.2 Kernel Regression Smoothing
6.3 Local Regression
6.4 Classes of Restricted Estimators
6.4.1 Ridge Regression
6.4.2 Lasso
6.5 Spline Estimators
6.5.1 Base Splines
6.5.2 Smoothing Splines
6.6 Wavelet Estimators
6.6.1 Wavelet Base
6.6.2 Wavelet Smoothing
6.7 Choosing the Smoothing Parameter
6.8 Bootstrap in Regression
6.9 Problems
References
Index
