Dear all,
You are cordially invited to attend the fifth webinar of the Fall 2023 CPGIS Educational Webinar Series - GeoAI-Driven Spatiotemporal Process Understanding And Forecasting. Please find attached the flyer.
Registration: https://www.eventbrite.com/e/2023-fall-cpgis-education-webinar-series-all-events-pass-tickets-718334496817
US Eastern Time: Thursday, 10/26/2023, 9:00-10:00 PM
Beijing Time: Friday, 10/27/2023, 9:00-10:00 AM
Guest Speaker: Dr. Kang Liu
Title: Knowledge-driven GeoAI Models for Human Trajectory Generation
Abstract: Human mobility data play a crucial role in many fields such as infectious diseases, transportation, and public safety. Although the development of Information and Communication Technologies (ICT) has made it easy to collect individual-level positioning records, raw individual trajectory data are still limited in availability and usability due to privacy issues. Developing models to generate synthetic trajectories that are statistically close to the real data is a promising solution. Existing trajectory generation methods can be divided into two groups: mechanistic models and machine learning models. Although both types of models have existed for many years, they are still developing independently within their respective domains without integration and complementation. Here we are interested in how to adaptively combine and leverage the strengths of both machine learning and mechanism models, and two knowledge-driven GeoAI models for human trajectory generation will be introduced.