Spatial analytics for urban modeling: the case of spatial interaction models
Date: Wednesday, May 26th, 2021 [12:00 PM – 1:00 PM]
Presentation by: Prof. Taylor Oshan, Dept. of Geographical Sciences
Abstract:
There is a recent surge of “new” areas of research focusing on developing and applying quantitative methods to study urban patterns and processes, such as urban analytics, urban informatics, and urban science. Though many of the tools and techniques associated with these emerging themes incorporate spatial data and concepts, there is a long history of spatial analysis methods being developed and applied for urban modeling.
In this talk, I highlight the trajectory of one particular class of spatial analysis technique – that of spatial interaction modeling – which focuses on explaining and predicting aggregate movements or flows between a
set of locations. This is done by providing a historical background and describing a few major milestones before then shifting focus to discuss some opportunities and challenges for this longstanding framework
within the modern data economy. Along the way I incorporate examples from my own work and finish with a few ideas to potentially enhance spatial interaction models of transportation and urban mobility.
Presentor Biography:
Prof. Taylor Oshan is an Assistant Professor in the Department of Geographical Sciences in the College of Behavioral and Social Sciences at UMD. He is also a faculty member at the Center for Geospatial Information
Science. He is broadly interested in characterizing spatial patterns and processes through the use of quantitative geographic methods, which typically falls under the banners of spatial analysis and spatial statistics, geographic information science, and the emerging discipline of spatial data science. His work typically involves modeling human processes within urban environments and therefore also intersects with the disciplines of computational social science, and urban informatics. Overall, my research has targeted the development of multivariate spatial statistics and how they can be used to capture how relationships change by spatial and temporal contexts. This includes issues of theory, interpretation, scalability, and integration of traditional geographic models with novel “big” datasets, as well as applications in public health, crime, urban mobility, and transportation systems. He also participates in open source software development to support my research and facilitate the replicability and reproducibility of the spatial sciences. He received his PhD in Geography from Arizona State University and his MA in Geography from CUNY Hunter College.