2022Journal article

Dynamic Population Mapping with AutoGluon

Song, Y., Xu, Y., Chen, B., He, Q., Tu, Y., Wang, F., & Cai, C.

Urban Informatics, 1(1), 13

Read publication ↗

Research question

Can automated machine learning improve scalable dynamic population mapping?

Geographic scope

China · dynamic gridded population

Evidence base

Automated machine learning and multi-source spatial data

Approach

AutoGluon is used to compare and ensemble models linking population signals with a wide range of physical and social geographic predictors.

Why it matters

The work reduces dependence on a single manually selected model and provides a reproducible route toward dynamic population surfaces.

Research connections

Where this paper sits in the lab’s research system