The Summer Training Workshop on Spatiotemporal Innovation 2024
The Summer Training Workshop on Spatiotemporal Innovation 2024
1730 Cambridge Street, Cambridge MA 02138, Cambridge, MA
July 16-19, 2024
Watching the Testimonials for 2023 Summer Workshop on Youtube
Sponsored by the Spatial Data Lab*, this hands-on workshop is to promote replicable and expandable spatiotemporal science with advanced methodology and technology. With a focus on geospatial analytics and GeoAI analytics, the workshop will discuss methodology, functions, workflow-based tools for spatiotemporal data analysis as well as case studies for their applications across different fields, including public health, business, social media, remote sensing, and environment. The workshop will offer an excellent opportunity for participants to network, collaborate, and further develop their leadership knowledge and skills.
Topics (4 days)
I: Visual Programming for Data Science with KNIME
1. An overview of GeoAI and replicable spatiotemporal data analysis
2. An Introduction to KNIME
- Basic functions of KNIME
- Open data access with KNIME
- Online Data Analysis and Visualization
3. Hands-on practice: Build a prototype workflow with KNIME
II: Spatial Data Analysis with KNIME Extensions
4. Geospatial Analytics for KNIME
- Nodes for spatial analysis
- Nodes for spatial data visualization
- Spatial Modelling with KNIME
5. Exploratory Spatial Data Analysis
- An overview of ESDA
- Spatial Weights
- Spatial Cluster Analysis (DBSCAN, SOM, SKATER, REDCAP, etc.)
- Spatial Autocorrelation
- Spatial Interpolation
- Hot Spot Analysis
- Spatial Dimensionality Reduction (GWPCA, GW Kernel PCA, Multidimensional Scaling with Spatial Constraints)
6. Spatial Autoregressive Models
- OLS with Spatial Test
- Spatial Autoregressive Models (SLM, SEM, SDM)
- Spatial Varying Models
- Multilevel Spatial Models
- Generalized Spatial Two/Three-Stage Least Squares
- Spatial Panel Data Models
7. Hands-on practice: Implement ESDA and Spatial Statistics in KNIME
III: Introduction to AI Models
8. Introduction to Basic AI Models in KNIME
- Introduction to AI Models: Machine Learning and Deep Learning
- Linear Regression Learner
- Logistic Regression Learner
- Decision Tree for Classification and Regression
- Text processing
9. Advanced Machine Learning Models in KNIME
- Support Vector Machine(SVM)-model
- Random Forest
- XGBoost Tree Ensemble Learner
- XGBoost Linear Model Learner
- Multiple Layer Perceptron (Neural Network)
10. Advanced Deep Learning Models
- Configuring Python environment for Deep Learning models in KNIME
- Convolutional Neural Network
- Long Short-Term Memory
- Graph Neural Network
11. Hands-on practice: Evaluate the Performance of Machine Learning Models
IV: Introduction to GeoAI models
12. Machine Learning in GeoAI Models
- Configuring R environment
- GeoAI: Spatially Explicit Artificial Intelligence
- Geographically Weighted AI Models (GW Random Forest, GW Support Vector Regression)
13. Deep Learning in Advanced GeoAI Models
- GW Extreme Learning Model(GWELM)
- GW Artificial Neural Network(GWANN)
- Spatial Autoencoders
- Spatial Embeddings
- Reinforcement Learning
- Deep Neural Network Models (GW Convolutional Neural Network, Graph Neural Network for Spatial Weight)
14. Hands-on practice: An Exemplary Case Study using GRF and GWANN
V: GeoAI Applications
- GeoAI for Healthcare Accessibility: Evaluating Commercial Dominants of Public Health with GRF
- GeoAI for Service Area Delimitation: Hospital Service Area Delimitation with GNN
- GeoAI for Remote Sensing:
- Introduction to KNIME GeoImage extension: Carbon Density Estimates
- GeoAI for Land Use /Land Cover: Classification techniques in LULC
- GeoAI for Social Media: Twitter Sentiment Analysis with KNIME
VI. An Integration of Tools for Spatial Data Analysis based on KNIME Platform
- KNIME + QGIS, ArcGIS, STATA, Google Earth Engine（Geemap）
- KNIME Business Hub
Requirement:
It is desirable that applicants have a background in geographic analysis. All participants are expected to complete a group project and make a presentation on their group’s research. Those who complete the program and participate in a group project will receive a certificate. Those outstanding participants will be invited to join the research team of the Spatial Data Lab project.
Application:
To apply, please submit your application, including your CV and the abstract of your research via this form: https://harvard.az1.qualtrics.com/jfe/form/SV_2aFjGNAa1AVcFQG before April 1, 2024. Application will be open until all seats are filled. Detailed agenda and lodging information will be sent to the accepted applicants later. Participants are responsible for their own travel and lodging expenses. Please visit http://spatialdatalab.org for more information or contact spatialdatalab@lists.fas.harvard.edu for questions.
Registration Fee:
• $2,980 registered/paid before 0:00 am ET, May 1, 2024
• $3,680 registered/paid after 0:00 am ET, May 1, 2024
Location:
The onsite event will take place at 1730 Cambridge Street, Cambridge MA 02138 (map), Cambridge, MA.
Contact:
spatialdatalab@lists.fas.harvard.edu