An Introduction to Spatial Data Science with GeoDa Volume 2 Clustering Spatial Data (

Cover
Half Title
Title Page
Copyright Page
Dedication
Contents
List of Figures
Preface
	Acknowledgments
About the Author
1. Introduction
	1.1. Overview of Volume 2
	1.2. Sample Data Sets
I. Dimension Reduction
	2. Principal Component Analysis (PCA)
		2.1. Topics Covered
		2.2. Matrix Algebra Review
			2.2.1. Eigenvalues and eigenvectors
			2.2.2. Matrix decompositions
		2.3. Principal Components
			2.3.1. Implementation
			2.3.2. Interpretation
		2.4. Visualizing principal components
			2.4.1. Scatter plot
			2.4.2. Multivariate decomposition
		2.5. Spatializing Principal Components
			2.5.1. Principal component map
			2.5.2. Univariate cluster map
			2.5.3. Principal components as multivariate cluster maps
	3. Multidimensional Scaling (MDS)
		3.1. Topics Covered
		3.2. Classic Metric Scaling
			3.2.1. Mathematical Details
			3.2.2. Implementation
		3.3. SMACOF
			3.3.1. Mathematical Details
			3.3.2. Implementation
		3.4. Visualizing MDS
			3.4.1. MDS and Parallel Coordinate Plot
			3.4.2. MDS Scatter Plot with Categories
			3.4.3. 3-D MDS
		3.5. Spatializing MDS
			3.5.1. MDS and Map
			3.5.2. MDS Spatial Weights
			3.5.3. MDS Neighbor Match Test
			3.5.4. HDBSCAN and MDS
	4. Stochastic Neighbor Embedding (SNE)
		4.1. Topics Covered
		4.2. Basics of Information Theory
			4.2.1. Stochastic Neighbors
		4.3. t-SNE
			4.3.1. Cost Function and Optimization
			4.3.2. Large Data Applications (Barnes-Hut)
		4.4. Implementation
			4.4.1. Animation
			4.4.2. Tuning the Optimization
			4.4.3. Interpretation and Spatialization
		4.5. Comparing Distance Preserving Methods
			4.5.1. Comparing t-SNE Options
			4.5.2. Local Fit with Common Coverage Percentage
II. Classic Clustering
	5. Hierarchical Clustering Methods
		5.1. Topics Covered
		5.2. Dissimilarity
		5.3. Agglomerative Clustering
			5.3.1. Linkage and Updating Formula
			5.3.2. Dendrogram
		5.4. Implementation
			5.4.1. Variable Settings Dialog
			5.4.2. Ward’s method
			5.4.3. Single linkage
			5.4.4. Complete linkage
			5.4.5. Average linkage
			5.4.6. Sensitivity Analysis
	6. Partitioning Clustering Methods
		6.1. Topics Covered
		6.2. The K Means Algorithm
			6.2.1. Iterative Relocation
			6.2.2. The Choice of K
			6.2.3. K-means++
		6.3. Implementation
			6.3.1. Digression: Clustering with Dimension Reduction
			6.3.2. Cluster Parameters
			6.3.3. Cluster Results
			6.3.4. Options and Sensitivity Analysis
		6.4. Cluster Categories as Variables
			6.4.1. Conditional Box Plot
			6.4.2. Aggregation by Cluster
	7. Advanced Clustering Methods
		7.1. Topics Covered
		7.2. K-Medians
			7.2.1. Implementation
			7.2.2. Options and Sensitivity Analysis
		7.3. K-Medoids
			7.3.1. The PAM Algorithm for K-Medoids
			7.3.2. Improving on the PAM Algorithm
			7.3.3. Implementation
			7.3.4. Options and Sensitivity Analysis
	8. Spectral Clustering
		8.1. Topics Covered
		8.2. Spectral Clustering Logic
		8.3. Clustering as a Graph Partitioning Problem
			8.3.1. Graph Laplacian
		8.4. The Spectral Clustering Algorithm
			8.4.1. Adjacency matrix
			8.4.2. Clustering on the Eigenvectors of the Graph Laplacian
			8.4.3. Spectral Clustering Parameters
		8.5. Implementation
			8.5.1. Cluster results
			8.5.2. Options and Sensitivity Analysis
III. Spatial Clustering
	9. Spatializing Classic Clustering Methods
		9.1. Topics Covered
		9.2. Clustering on Geographic Coordinates
			9.2.1. Implementation
		9.3. Including Geographical Coordinates in the Feature Set
			9.3.1. Implementation
		9.4. Weighted Optimization of Geographical and Attribute Similarity
			9.4.1. Optimization
			9.4.2. Implementation
		9.5. Constructing a Spatially Contiguous Solution
			9.5.1. Implementation
	10. Spatially Constrained Clustering – Hierarchical Methods
		10.1. Topics Covered
		10.2. Spatially Constrained Hierarchical Clustering (SCHC)
			10.2.1. The Algorithm
			10.2.2. Implementation
		10.3. SKATER
			10.3.1. Pruning the Minimum Spanning Tree
			10.3.2. Implementation
		10.4. REDCAP
			10.4.1. Illustration – FullOrder-CompleteLinkage
			10.4.2. Implementation
		10.5. Assessment
	11. Spatially Constrained Clustering – Partitioning Methods
		11.1. Topics Covered
		11.2. Automatic Zoning Procedure (AZP)
			11.2.1. AZP Heuristic
			11.2.2. Tabu Search
			11.2.3. Simulated Annealing
			11.2.4. ARiSeL
			11.2.5. Using the Outcome from Another Cluster Routine as the Initial Feasible Region
			11.2.6. Implementation
			11.2.7. Search Options
			11.2.8. Initialization Options
		11.3. Max-P Region Problem
			11.3.1. Max-p Heuristic
			11.3.2. Implementation
			11.3.3. Sensitivity Analysis
IV. Assessment
	12. Cluster Validation
		12.1. Topics Covered
		12.2. Internal Validity
			12.2.1. Traditional Measures of Fit
			12.2.2. Balance
			12.2.3. Join Count Ratio
			12.2.4. Compactness
			12.2.5. Connectedness
			12.2.6. Implementation
		12.3. External Validity
			12.3.1. Classic Measures
			12.3.2. Visualizing Cluster Match
		12.4. Beyond Clustering
Bibliography
Index