Available:*
Library | Item Barcode | Call Number | Material Type | Item Category 1 | Status |
---|---|---|---|---|---|
Searching... | 30000010029168 | QA278 L39 2002 | Open Access Book | Book | Searching... |
On Order
Summary
Summary
Research has generated a number of advances in methods for spatial cluster modelling in recent years, particularly in the area of Bayesian cluster modelling. Along with these advances has come an explosion of interest in the potential applications of this work, especially in epidemiology and genome research.
In one integrated volume, this book reviews the state-of-the-art in spatial clustering and spatial cluster modelling, bringing together research and applications previously scattered throughout the literature. It begins with an overview of the field, then presents a series of chapters that illuminate the nature and purpose of cluster modelling within different application areas, including astrophysics, epidemiology, ecology, and imaging. The focus then shifts to methods, with discussions on point and object process modelling, perfect sampling of cluster processes, partitioning in space and space-time, spatial and spatio-temporal process modelling, nonparametric methods for clustering, and spatio-temporal cluster modelling.
Many figures, some in full color, complement the text, and a single section of references cited makes it easy to locate source material. Leading specialists in the field of cluster modelling authored each chapter, and an introduction by the editors to each chapter provides a cohesion not typically found in contributed works. Spatial Cluster Modelling thus offers a singular opportunity to explore this exciting new field, understand its techniques, and apply them in your own research.
Author Notes
Lawson\, Andrew B.; Denison\, David G.T.
Table of Contents
List of Contributors | p. xi |
Preface | p. xiii |
1 Spatial Cluster Modelling: An Overview | p. 1 |
1.1 Introduction | p. 1 |
1.2 Historical Development | p. 3 |
1.2.1 Conventional Clustering | p. 6 |
1.2.2 Spatial Clustering | p. 8 |
1.3 Notation and Model Development | p. 12 |
1.3.1 Nonparametric Approaches | p. 13 |
1.3.2 Point or Object Process Modelling | p. 15 |
1.3.3 Random Effect Modelling | p. 16 |
1.3.4 Partition Modelling | p. 18 |
1.3.5 Spatio-Temporal Process Modelling | p. 18 |
I Point process cluster modelling | p. 21 |
2 Significance in Scale-Space for Clustering | p. 23 |
2.1 Introduction | p. 23 |
2.2 Overview | p. 24 |
2.3 New Method | p. 29 |
2.4 Future Directions | p. 35 |
3 Statistical Inference for Cox Processes | p. 37 |
3.1 Introduction | p. 37 |
3.2 Poisson Processes | p. 39 |
3.3 Cox Processes | p. 41 |
3.4 Summary Statistics | p. 43 |
3.5 Parametric Models of Cox Processes | p. 45 |
3.5.1 Neyman-Scott Processes as Cox Processes | p. 45 |
3.5.2 Log Gaussian Cox Processes | p. 48 |
3.5.3 Shot Noise G Cox Processes | p. 49 |
3.6 Estimation for Parametric Models of Cox Processes | p. 51 |
3.7 Prediction | p. 54 |
3.7.1 Conditional Simulation for Neyman-Scott Processes | p. 55 |
3.7.2 Conditional Simulation for LGCPs | p. 55 |
3.7.3 Conditional Simulation for Shot-noise G Cox Processes | p. 56 |
3.8 Discussion | p. 58 |
4 Extrapolating and Interpolating Spatial Patterns | p. 61 |
4.1 Introduction | p. 61 |
4.2 Formulation and Notation | p. 62 |
4.2.1 Germ-grain Models | p. 63 |
4.2.2 Problem Statement | p. 63 |
4.2.3 Edge Effects and Sampling Bias | p. 64 |
4.2.4 Extrapolation | p. 65 |
4.3 Spatial Cluster Processes | p. 66 |
4.3.1 Independent Cluster Processes | p. 67 |
4.3.2 Cox Cluster Processes | p. 68 |
4.3.3 Cluster Formation Densities | p. 69 |
4.4 Bayesian Cluster Analysis | p. 72 |
4.4.1 Markov Point Processes | p. 72 |
4.4.2 Sampling Bias for Independent Cluster Processes | p. 74 |
4.4.3 Spatial Birth-and-Death Processes | p. 75 |
Example: Redwood Seedlings | p. 76 |
4.4.4 Parameter Estimation | p. 78 |
Example: Cox-Matern Cluster Process | p. 80 |
4.4.5 Adaptive Coupling from the Past | p. 80 |
4.4.6 Example: Cox-Matern Cluster Process | p. 84 |
4.5 Summary and Conclusion | p. 86 |
5 Perfect Sampling for Point Process Cluster Modelling | p. 87 |
5.1 Introduction | p. 87 |
5.2 Bayesian Cluster Model | p. 89 |
5.2.1 Preliminaries | p. 89 |
5.2.2 Model Specification | p. 90 |
5.3 Sampling from the Posterior | p. 93 |
5.4 Specialized Examples | p. 95 |
5.4.1 Neyman-Scott Model | p. 95 |
5.4.2 Pure Silhouette Models | p. 98 |
5.5 Leukemia Incidence in Upstate New York | p. 100 |
5.6 Redwood Seedlings Data | p. 106 |
6 Bayesian Estimation and Segmentation of Spatial Point Processes Using Voronoi Tilings | p. 109 |
6.1 Introduction | p. 109 |
6.2 Proposed Solution Framework | p. 110 |
6.2.1 Formulation | p. 110 |
6.2.2 Voronoi Tilings | p. 111 |
6.2.3 Markov Chain Monte Carlo Using Dynamic Voronoi Tilings | p. 111 |
6.3 Intensity Estimation | p. 112 |
6.3.1 Formulation | p. 112 |
6.3.2 MCMC Implementation | p. 112 |
Fixed Number of Tiles | p. 113 |
Variable Number of Tiles | p. 114 |
6.4 Intensity Segmentation | p. 115 |
6.4.1 Formulation | p. 115 |
Fixed Number of Tiles | p. 115 |
Variable Number of Tiles | p. 117 |
6.5 Examples | p. 117 |
6.5.1 Simulated Examples | p. 117 |
A Sine Wave | p. 117 |
A Linear Feature | p. 118 |
6.5.2 New Madrid Seismic Region | p. 118 |
6.6 Discussion | p. 119 |
II Spatial process cluster modelling | p. 123 |
7 Partition Modelling | p. 125 |
7.1 Introduction | p. 125 |
7.2 Partition Models | p. 126 |
7.2.1 Partitioning for Spatial Data | p. 127 |
7.2.2 Bayesian Inference | p. 128 |
7.2.3 Predictive Inference | p. 131 |
7.2.4 Markov Chain Monte Carlo Simulation | p. 132 |
7.2.5 Partition Model Prior | p. 133 |
7.3 Piazza Road Dataset | p. 135 |
7.4 Spatial Count Data | p. 135 |
7.4.1 The Poisson-Gamma Model for Disease Mapping | p. 137 |
7.4.2 Disease Mapping with Covariates | p. 138 |
7.4.3 East German Lip Cancer Dataset | p. 141 |
7.5 Discussion | p. 144 |
7.6 Further Reading | p. 144 |
8 Cluster Modelling for Disease Rate Mapping | p. 147 |
8.1 Introduction | p. 147 |
8.2 Statistical Model | p. 148 |
8.3 Posterior Calculation | p. 150 |
8.4 Example: U.S. Cancer Mortality Atlas | p. 153 |
8.4.1 Breast Cancer | p. 154 |
8.4.2 Cervical Cancer | p. 155 |
8.4.3 Colorectal Cancer | p. 155 |
8.4.4 Lung Cancer | p. 159 |
8.4.5 Stomach Cancer | p. 159 |
8.5 Conclusions | p. 159 |
9 Analyzing Spatial Data Using Skew-Gaussian Processes | p. 163 |
9.1 Introduction | p. 163 |
9.2 Skew-Gaussian Processes | p. 164 |
9.2.1 The Model | p. 165 |
9.2.2 Bayesian Analysis | p. 166 |
9.2.3 Computational Strategy | p. 167 |
9.3 Real Data Illustration: Spatial Potential Data Prediction | p. 168 |
9.4 Discussion | p. 171 |
10 Accounting for Absorption Lines in Images Obtained with the Chandra X-ray Observatory | p. 175 |
10.1 Statistical Challenges of the Chandra X-ray Observatory | p. 175 |
10.2 Modeling the Image | p. 179 |
10.2.1 Model-Based Spatial Analysis | p. 179 |
10.2.2 Model-Based Spectral Analysis | p. 183 |
10.3 Absorption Lines | p. 184 |
10.3.1 Scientific Background | p. 184 |
10.3.2 Statistical Models | p. 185 |
10.3.3 Model Fitting | p. 188 |
10.3.4 A Simulation-Based Example | p. 190 |
10.4 Spectral Models with Absorption Lines | p. 192 |
10.4.1 Combining Models and Algorithms | p. 192 |
10.4.2 An Example | p. 195 |
10.5 Discussion | p. 196 |
11 Spatial Modelling of Count Data: A Case Study in Modelling Breeding Bird Survey Data on Large Spatial Domains | p. 199 |
11.1 Introduction | p. 199 |
11.2 The Poisson Random Effects Model | p. 200 |
11.2.1 Spectral Formulation | p. 202 |
11.2.2 Model Implementation and Prediction | p. 204 |
Selected Full-Conditional Distributions | p. 205 |
Prediction | p. 206 |
Implementation | p. 206 |
11.3 Results | p. 206 |
11.4 Conclusion | p. 207 |
III Spatio-temporal cluster modelling | p. 211 |
12 Modelling Strategies for Spatial-Temporal Data | p. 213 |
12.1 Introduction | p. 213 |
12.2 Modelling Strategy | p. 214 |
12.3 D-D (Drift-Drift) Models | p. 215 |
12.4 D-C (Drift-Correlation) Models | p. 220 |
12.5 C-C (Correlation-Correlation) Models | p. 222 |
12.6 A Unified Analysis on the Circle | p. 224 |
12.7 Discussion | p. 225 |
13 Spatio-Temporal Partition Modelling: An Example from Neurophysiology | p. 227 |
13.1 Introduction | p. 227 |
13.2 The Neurophysiological Experiment | p. 227 |
13.3 The Linear Inverse Solution | p. 228 |
13.4 The Mixture Model | p. 229 |
13.4.1 Initial Preparation of the Data | p. 229 |
13.4.2 Formulation of the Mixture Model | p. 230 |
13.5 Classification of the Inverse Solution | p. 232 |
13.6 Discussion | p. 234 |
14 Spatio-Temporal Cluster Modelling of Small Area Health Data | p. 235 |
14.1 Introduction | p. 235 |
14.2 Basic Cluster Modelling approaches | p. 235 |
14.2.1 Case Event Data Models | p. 236 |
Inter-Event versus Hidden Process Modelling | p. 236 |
14.2.2 Small Area Count Data Models | p. 239 |
14.2.3 Spatio-Temporal Extensions to Cluster Models | p. 239 |
14.3 A Spatio-Temporal Hidden Process Model | p. 240 |
14.4 Model Development | p. 240 |
14.4.1 Estimation of g(s,t) | p. 243 |
The Prior Distribution for Cluster Centres | p. 244 |
Choice of Cluster Distribution Function | p. 245 |
Other Prior Distributions and the Posterior Distribution | p. 245 |
14.4.2 Region Counts | p. 246 |
Integrated Intensity | p. 247 |
14.5 The Posterior Sampling Algorithm | p. 248 |
14.5.1 Goodness-of-Fit Measures for the Model | p. 249 |
14.6 Data Example: Scottish Birth Abnormalities | p. 249 |
14.6.1 Introduction | p. 249 |
14.6.2 Exploratory Analysis | p. 250 |
14.6.3 Model Fitting Results | p. 252 |
14.7 Discussion | p. 256 |
References | p. 259 |
Index | p. 277 |
Author Index | p. 281 |