Deep Learning of Geospatial Patterns & Applications II
Session: GeoAI Symposium: Deep Learning of Geospatial Patterns & Applications II
Type: Paper
Sponsor Groups: Geographic Information Science and Systems Specialty Group, Spatial Analysis and Modeling Specialty Group, Cyberinfrastructure Specialty Group
Organizers: Di Zhu, Ximeng Cheng, Yu Liu
Chairs: Di Zhu
Call for Submissions
In this session within the GeoAI symposiums, we seek innovative analysis methods and applications that advance our traditional understanding of geographic/spatial phenomena at urban\regional\global scales. The session highlights the development of cutting-edge geospatial modeling techniques that are rooted in geography theories and the novel deep/machine learning ideas that are of spatial thinking. Potential topics include, but are not limited to:
- The combination of deep learning models and traditional spatial analysis/statistics methods.
- Understanding urban settlements from various aspects (e.g. population, urban built-up area, expansion&evolution, socio-economic properties).
- Spatial network analysis and graph-based neural network.
- Generating continuous geographic field patterns based on sparse observations.
- Spatial predictability of unknown attributes.
- Intra-urban/regional applications that are based on the geographic knowledge uncovered by deep/machine learning models.
Organizers:
Di Zhu (patrick.zhu@pku.edu.cn), Peking University
Ximeng Cheng (chengximeng@pku.edu.cn), Peking University
Yu Liu (liuyu@urban.pku.edu.cn), Peking University
If you are interested in this session, please send your abstract and the Personal Identification Number (PIN) for AAG 2020 to Di Zhu (patrick.zhu@pku.edu.cn) or Ximeng Cheng (chengximeng@pku.edu.cn) or Yu Liu (liuyu@urban.pku.edu.cn) by October 31, 2019.
Descriptions
Deep learning approaches have been increasingly used to understand spatial processes from a data-driven perspective, as they have been proven to be efficient for high-dimensional data representation and complex function approximation. Geographers utilize the deep learning techniques in understanding geospatial patterns, innovative research has been proposed in areas including object detection, spatial interpolation, spatial network analysis, pattern classification, spatial optimization and so on. However, considering "spatial" as "special", the nature of geospatial patterns is different from patterns in other fields. Thus, specific spatial thinking on applying deep learning is necessary to understand complex geographic patterns.
We hope to gather innovative researchers with novel analysis methods and applications that advance our traditional understanding of geographic/spatial phenomena at urban\regional\global scales. The session highlights the discussion of cutting-edge geospatial modeling techniques that are rooted in geography theories and the novel deep/machine learning ideas that are of spatial thinking.
If you are interested in this session, please send your abstract and the Personal Identification Number (PIN) for AAG 2020 to Di Zhu (patrick.zhu@pku.edu.cn), Ximeng Cheng (chengximeng@pku.edu.cn) and Yu Liu (liuyu@urban.pku.edu.cn) by October 31, 2019.
素材来源：S?-Lab
材料整理：朱递
内容排版：杨桃汝