Detect geospatial boundaries of communities, describe dynamic community profiles, and identify key transitions within our community structure during the COVID-19 pandemic within the Twin Cities Metro Area.
Explore the frontier of Geospatial Artificial Intelligence, and bridge the methodological linkage between deep/machine learning models and spatial analytical models with a focus on human-environment complexities within socioeconomic and population data.
Collaboration with Minnesota Population Center
Investigate systematic methods for analyzing multi-modal spatial networks at different spatio-temporal scales. Develop a WebGIS platform and apply to: city (Shenzhen), megalopolis (Guangdong-Hong Kong- Macau Big Bay Area), and nation (China).
Investigate human behavior characteristics from the perspective of the interaction between people and geographical environment with the support of big geo-data using deep learning methods using PyTorch on Linux.
Represent and model diverse geospatial semantics of locations and develop spatial prediction approaches incorporating locations' relatedness.
Develop a spatio-temporal Geo-propagation method for sparse geospatial data prediction with an application of the house price estimation in Beijing from 2011-2018.
Integrate existing algorithms of location-related schedule planning and location recommendation in the context of clients' business scenarios: negotiate time according to every participant's schedule and activity preference.
Investigate the GPS-enabled taxis' origin and destination (OD) distributions, mobility patterns and relations with urban structure, street networks. Develop spatio-temporal data mining algorithms for processing large-scale geo-data using Python and PostgreSQL.