Urban computing / Yu Zheng.

Includes bibliographical references and index.

1.Overview
1.1.Introduction
1.2.Definition of Urban Computing
1.3.General Framework
1.3.1.Brief and Example
1.3.2.Functions of Each Layer
1.4.Key Urban-Computing Challenges
1.4.1.Urban-Sensing Challenges
1.4.2.Urban Data Management Challenges
1.4.3.Urban Data Analytics Challenges
1.4.4.Urban Service Challenges
1.5.Urban Data
1.5.1.Taxonomy of Urban Data
1.5.2.Geographical Data
1.5.3.Traffic Data on Road Networks
1.5.4.Mobile Phone Data
1.5.5.Commuting Data
1.5.6.Environmental-Monitoring Data
1.5.7.Social Network Data
1.5.8.Energy
1.5.9.Economy
1.5.10.Health Care
1.6.Public Datasets
References
2.Urban-Computing Applications
2.1.Introduction
2.2.Urban Computing for Urban Planning
2.2.1.Gleaning Underlying Problems in Transportation Networks
2.2.2.Discovering Functional Regions
2.2.3.Detecting a City's Boundaries
2.2.4.Facility and Resource Deployment
Contents note continued: 2.3.Urban Computing for Transportation Systems
2.3.1.Improving Driving Experiences
2.3.2.Improving Taxi Services
2.3.3.Improving Bus Services
2.3.4.Subway Services
2.3.5.Bike-Sharing Systems
2.4.Urban Computing for the Environment
2.4.1.Air Quality
2.4.2.Noise Pollution
2.4.3.Urban Water
2.5.Urban Computing for Urban Energy Consumption
2.5.1.Gas Consumption
2.5.2.Electricity Consumption
2.6.Urban Computing for Social Applications
2.6.1.Concepts of Location-Based Social Networks
2.6.2.Understanding Users in Location-Based Social Networks
2.6.3.Location Recommendations
2.7.Urban Computing for the Economy
2.7.1.Location Selection for Businesses
2.7.2.Optimizing Urban Logistics
2.8.Urban Computing for Public Safety and Security
2.8.1.Detecting Urban Anomalies
2.8.2.Predicting the Flow of Crowds
2.9.Summary
3.Urban Sensing
3.1.Introduction
Contents note continued: 3.1.1.Four Paradigms of Urban Sensing
3.1.2.General Framework of Urban Sensing
3.2.Sensor and Facility Deployment
3.2.1.Finding Optimal Meeting Points
3.2.2.Maximizing Coverage
3.2.3.Learning-to-Rank Candidates
3.2.4.Minimizing Uncertainty
3.3.Human-Centric Urban Sensing
3.3.1.Data Evaluation
3.3.2.Participant Recruitment and Task Design
3.4.Filling Missing Values
3.4.1.Problem and Challenges
3.4.2.Spatial Models
3.4.3.Temporal Models
3.4.4.Spatiotemporal Models
3.5.Summary
4.Spatiotemporal Data Management
4.1.Introduction
4.1.1.Data Structures
4.1.2.Queries
4.1.3.Indexes
4.1.4.Retrieval Algorithms
4.2.Data Structures
4.2.1.Point-Based Spatial Static Data
4.2.2.Point-Based Spatial Time Series Data
4.2.3.Point-Based Spatiotemporal Data
4.2.4.Network-Based Spatial Static Data
4.2.5.Network-Based Spatial Time Series Data
Contents note continued: 4.2.6.Network-Based Spatiotemporal Data
4.3.Spatial Data Management
4.3.1.Grid-Based Spatial Index
4.3.2.Quadtree-Based Spatial Index
4.3.3.K-D Tree-Based Spatial Index
4.3.4.R-Tree-Based Spatial Index
4.4.Spatiotemporal Data Management
4.4.1.Managing Spatial Static and Temporal Dynamic Data
4.4.2.Moving-Object Databases
4.4.3.Trajectory Data Management
4.5.Hybrid Indexes for Managing Multiple Datasets
4.5.1.Queries and Motivations
4.5.2.Spatial Key Words
4.5.3.Indexes for Managing Multiple Datasets
4.6.Summary
5.Introduction to Cloud Computing
5.1.Introduction
5.2.Storage
5.2.1.SQL Databases
5.2.2.Azure Storage
5.2.3.Redis Cache
5.3.Computing
5.3.1.Virtual Machine
5.3.2.Cloud Services
5.3.3.HDInsight
5.4.Applications
5.4.1.Web Apps
5.4.2.Mobile Apps
5.4.3.API Apps
5.5.Summary
6.Managing Spatiotemporal Data in the Cloud
Contents note continued: 6.1.Introduction
6.1.1.Challenges
6.1.2.General Data Management Schemes on the Cloud
6.2.Managing Point-Based Data
6.2.1.Managing Point-Based Spatiotemporal Static Data
6.2.2.Managing Point-Based Spatial Static and Temporal Dynamic Data
6.2.3.Managing Point-Based Spatiotemporal Dynamic Data
6.3.Managing Network-Based Data
6.3.1.Managing Spatiotemporal Static Networks
6.3.2.Managing Network-Based Spatial Static and Temporally Dynamic Data
6.3.3.Managing Network-Based Spatiotemporal Dynamic Data
6.4.Urban Big-Data Platform
6.5.Summary
7.Fundamental Data-Mining Techniques for Urban Data
7.1.Introduction
7.1.1.General Framework of Data Mining
7.1.2.Relationship between Data Mining and Related Technologies
7.2.Data Preprocessing
7.2.1.Data Cleaning
7.2.2.Data Transformation
7.2.3.Data Integration
7.3.Frequent Pattern Mining and Association Rules
7.3.1.Basic Concepts
Contents note continued: 7.3.2.Frequent Itemset Mining Methods
7.3.3.Sequential Pattern Mining
7.3.4.Frequent Subgraph Pattern Mining
7.4.Clustering
7.4.1.Concepts
7.4.2.Partitioning Clustering Methods
7.4.3.Density-Based Clustering
7.4.4.Hierarchical Clustering Methods
7.5.Classification
7.5.1.Concepts
7.5.2.Naive Bayesian Classification
7.5.3.Decision Trees
7.5.4.Support Vector Machines
7.5.5.Classification with Imbalanced Data
7.6.Regression
7.6.1.Linear Regression
7.6.2.Autoregression
7.6.3.Regression Tree
7.7.Outlier and Anomaly Detection
7.7.1.Proximity-Based Outlier Detection
7.7.2.Statistic-Based Outlier Detection
7.8.Summary
8.Advanced Machine-Learning Techniques for Spatiotemporal Data
8.1.Introduction
8.2.Unique Properties of Spatiotemporal Data
8.2.1.Spatial Properties of Spatiotemporal Data
8.2.2.Temporal Properties
8.3.Collaborative Filtering
Contents note continued: 8.3.1.Basic Models: User Based and Item Based
8.3.2.Collaborative Filtering for Spatiotemporal Data
8.4.Matrix Factorization
8.4.1.Basic Matrix Factorization Methods
8.4.2.Matrix Factorization for Spatiotemporal Data
8.5.Tensor Decomposition
8.5.1.Basic Concepts of Tensors
8.5.2.Methods of Tensor Decomposition
8.5.3.Tensor Decomposition for Spatiotemporal Data
8.6.Probabilistic Graphical Models
8.6.1.General Concepts
8.6.2.Bayesian Networks
8.6.3.Markov Random Field
8.6.4.Bayesian Networks for Spatiotemporal Data
8.6.5.Markov Networks for Spatiotemporal Data
8.7.Deep Learning
8.7.1.Artificial Neural Networks
8.7.2.Convolutional Neural Networks
8.7.3.Recurrent Neural Networks
8.7.4.Deep Learning for Spatiotemporal Data
8.8.Reinforcement Learning
8.8.1.Concepts of Reinforcement Learning
8.8.2.Tabular Action-Value Methods
8.8.3.Approximate Methods
8.9.Summary
Contents note continued: 9.Cross-Domain Knowledge Fusion
9.1.Introduction
9.1.1.Relationship to Traditional Data Integration
9.1.2.Relationship to Heterogeneous Information Networks
9.2.Stage-Based Knowledge Fusion
9.3.Feature-Based Knowledge Fusion
9.3.1.Feature Concatenation with Regularization
9.3.2.Deep Learning-Based Knowledge Fusion
9.4.Semantic Meaning-Based Knowledge Fusion
9.4.1.Multi-View-Based Knowledge Fusion
9.4.2.Similarity-Based Knowledge Fusion
9.4.3.Probabilistic Dependency-Based Knowledge Fusion
9.4.4.Transfer Learning-Based Knowledge Fusion
9.5.Comparison between Different Fusion Methods
9.5.1.Volume, Properties, and Insight of Datasets
9.5.2.The Goal of a Machine-Learning Task
9.5.3.Learning Approach of Machine-Learning Algorithms
9.5.4.Efficiency and Scalability
9.6.Summary
10.Advanced Topics in Urban Data Analytics
10.1.How to Select Useful Datasets
Contents note continued: 10.1.1.Understanding Target Problems
10.1.2.Insights behind Data
10.1.3.Validating Assumptions
10.2.Trajectory Data Mining
10.2.1.Trajectory Data
10.2.2.Trajectory Preprocessing
10.2.3.Trajectory Data Management
10.2.4.Uncertainty in a Trajectory
10.2.5.Trajectory Pattern Mining
10.2.6.Trajectory Classification
10.2.7.Anomalies Detection from Trajectories
10.2.8.Transferring Trajectories to Other Representations
10.3.Combining Machine Learning with Data Management
10.3.1.Motivation
10.3.2.Boosting Machine Learning with Indexing Structures
10.3.3.Scale Down Candidates for Machine Learning
10.3.4.Derive Bounds to Prune Computing Spaces for Machine Learning
10.4.Interactive Visual Data Analytics
10.4.1.Incorporating Multiple Complex Factors
10.4.2.Adjusting Parameters without Prior Knowledge
10.4.3.Drilling Down into Results
10.5.Summary
References.

Although there are a few books on urban informatics, this is the first book dedicated to urban computing, with a broad spectrum of coverage and an authoritative overview. This book introduces a general framework, key research problems, methodologies and applications of urban computing from a computer science perspective. More specifically, this book focuses on data and computing, distinguishing urban computing from tradition urban science based on classical models and empirical assumptions. Rapid urbanization has led to the expansion of numerous large cities, not only modernizing many people's lives but also posing big challenges, such as air pollution, energy consumption and traffic congestion. Tackling these challenges seemed nearly impossible only a few years ago given the complex and dynamic settings of cities. Nowadays, sensing technologies and large-scale computing infrastructure have produced a variety of big data, such as human mobility, meteorology, traffic patterns and geographical data. The corresponding big data implies rich knowledge about a city and can help tackle these challenges when used correctly. In addition, the rise of computing technology, such as cloud computing and artificial intelligence (AI), has provided us with unprecedented data processing capabilities--