Symposium on Frontiers in CyberGIS and Geospatial Data Science: Spatial Machine Learning and Deep Learning

Type: Paper
Sponsor Groups: Cyberinfrastructure Specialty Group, Geographic Information Science and Systems Specialty Group, Spatial Analysis and Modeling Specialty Group
Organizers: Jeon-Young Kang, Jin Xing, Shaowen Wang
Chairs: Jin Xing

Call for Submissions

In this session, we are particularly interested in new methodologies, algorithms, and applications of spatial machine learning approaches to address various challenges in geospatial data science. Major topics may include but are not limited to:

1. Innovative machine learning algorithms for geospatial modeling and simulation;
2. Novel deep learning architectures for geospatial data science, such as Attention Mechanism and transfer learning;
3. Exploration of machine learning algorithms in smart cities;
4. Natural language processing for text-based geospatial information;
5. Computer vision methods for remote sensing image analysis, including land cover classification, change detection, image fusion, and geo-registration;
6. Using machine learning or deep learning for geospatial data collection, especially with UAV and IoT;
7. Transfer learning for geospatial data mining and interpretation;
8. The integration strategies of machine learning and CyberGIS.


The rapid growth of geospatial data has far exceeded our previous capability of data analytics. In addition to new geospatial computing technologies, GIScientists continuously explore new methods to shift geospatial data analytics towards automatic model building. Recent advancement in deep learning algorithms has proven its success in automatically learning the representative and discriminative features in a hierarchical manner from geospatial big data. For example, high accuracy of land cover classification map has been generated using various convolutional neural network with remotely sensed imagery datasets. However, geospatial data science poses unique challenges in machine learning, such as large-scale network analysis, spatial optimization with scale heterogeneity, multi-temporal modelling, and location inference from large text corpus, to name a few here.


Type Details Minutes
Presenter Behnam Nikparvar*, University of North Carolina - Charlotte, Jean-Claude Thill, University of North Carolina at Charlotte, A Probabilistic Principal Components Analysis (PPCA) Approach to Impute Missing Values in Spatiotemporal Datasets 15
Presenter Nattapon Jaroenchai*, University of Illinois Urbana-Champaign, Zewei Xu, University of Illinois at Urbana-Champaign, Shaowen Wang, University of Illinois at Urbana-Champaign, Hydrographic Streamline Detection Using LiDAR and Deep Learning 15
Presenter Matthew Tenney*, University of Toronto - Scarborough, Ontario, GIS 9000 | 2019: A GeoSpatial Odyssey 15
Presenter Zhengcong Yin*, Texas A&M University, Andong Ma, Texas A&M University, Daniel W. Goldberg, Texas A&M University, A Deep Learning Approach for Rooftop Geocoding 15

To access contact information login