Yuki (Yuqing) Wang

Hi, I am

Yuki (Yuqing) Wang.

PhD Candidate in Criminology, Law and Society at University of California, Irvine.

I model how offenders make locational choices and how routine activities shape spatial exposure to victimization risk.

Spatial and Temporal Analysis Network Measurement Computational Acceleration Discrete Choice Models
Publications Contact

About Me

I am a PhD candidate in Criminology, Law and Society at the University of California, Irvine, and a member of the Irvine Laboratory for the Study of Space and Crime (ILSSC). My research asks why offenders travel to some places to commit crime and how people's routine activities expose them to victimization risk. I approach these questions through discrete choice models, spatial and temporal analysis, and measures of relational activity recorded in police data.

On the computational side, I build memory-efficient estimation pipelines in JAX for large spatial-choice models. The software makes it possible to use the full offender and victim samples and evaluate every Dallas spatial alternative without sampling observations or locations, while controlling memory demands on CPU or GPU hardware. My work integrates administrative records, census demographics, and business establishment data to study crime patterns at multiple geographic scales.

I also serve as a reviewer for Social Science & Medicine, Crime Prevention and Community Safety, Machine Learning: Science and Technology, and the International Joint Conference on Neural Networks (IJCNN 2025).

Quick Links

Publications

Map of offender, victim, and incident location clusters across Dallas

Investigating How Social and Physical Distance Impact Offender and Victim Mobility with Discrete Choice Modeling

Yuqing Wang, John R. Hipp — Journal of Quantitative Criminology, 2026

Examines how physical distance and social distance shape where offenders commit crimes and where victims are exposed to crime risk in Dallas. Uses discrete choice models with police incident and arrest records from 2014 to 2020, covering burglary, larceny, vehicle theft, assault, robbery, and drug violations at the census block group level. Finds strong distance decay for both offenders and victims, with offenders generally less spatially constrained, especially for property crimes. Racial dissimilarity reduces both offender and victim mobility, while income difference constrains victim mobility more consistently than offender mobility.

View →
Map showing ¼ mile buffers around three random blocks in a census tract

Measuring the Spatial Scale of Structural Racism and Discrimination

John R. Hipp, Yuqing Wang, et al. — Social Science & Medicine, 2026

Explores how structural racism operates at different spatial scales and its consequences for life expectancy. Constructs novel meso-level measures — including eviction rates, housing vacancies, loan denials, and proximity to toxic waste — at four buffer radii, alongside county-level racial inequality. Finds a nonlinear relationship where life expectancy drops sharply at higher concentrations of multidimensional structural racism. My contribution: data curation (geocoding, multi-source spatial data processing) and geospatial concept visualization.

View →
Geographic distribution of census blocks in this study

How Does the Business Environment Shape Mobility by Offenders and Mobile Targets?

John R. Hipp, Yuqing Wang — Crime & Delinquency, 2026

Uses discrete choice models on more than 653,000 geocoded crime records to examine how business composition at the census-block level shapes offender mobility and the spatial exposure underlying observed victimization. Finds that blocks with a greater mix of consumer-facing businesses are especially attractive to offenders and that the surrounding quarter-mile business context further increases targeting. The JAX estimator uses the full sample and every census-block alternative without sampling observations or locations.

View →
Population Partition diagram showing deterrence regions by discount rate and wealth

The Optimal Deterrence of Crime: A Focus on the Time Preference of DWI Offenders

Yuqing Wang, Yan Ru Pei — arXiv, 2019

Develops a general model for finding optimal penal strategies based on offenders' behavioral traits. Uses hyperbolic time discounting and prospect theory with a 207-participant survey to empirically characterize discount-rate distributions (zero-inflated exponential), then maximizes a social welfare function balancing deterrence benefits against implementation costs for DWI offenses.

arXiv →

Contact

Email: yqwang1@uci.edu

Office: Social Ecology II, University of California, Irvine

Social: LinkedIn · GitHub · Google Scholar · ORCID