Selected Publications

Knowing which transportation modes citizens use is critical for smart cities and planners. We show that scanning pervasive Wi-Fi access points with mobile phones can enhance GPS and geographic information to improve transportation detection and identify public transportation modes while conserving battery. This approach yields an average F1 score of 85% for inferring five common modes. Wi-Fi features improve detection accuracy of bus trips, train travel, and driving compared to GPS features alone and can substitute for GIS features. We conclude that crowdsourced Wi-Fi has been underutilized in transportation research and can improve mobile travel surveys and urban sensing applications.
-, 2018

Identifying the factors which influence academic performance is an essential part of educational research. Previous studies have documented the importance of personality traits, class attendance, and social network structure. However, because most of these analyses were based on a single behavioral aspect and/or small sample sizes, we lack a quantification of the interplay of these factors. Here, we infer the academic performance among a cohort of almost 1,000 freshmen based on data collected through smartphones, which the students used as their primary phone for two years. The availability of multi-channel data from a single population allows us to directly compare the predictive power of individual and social characteristics. We find that a student’s performance is best inferred from their social ties. Network indicators out-perform models based on individual characteristics. We confirm earlier findings indicating that class attendance is the most important predictor among the individual characteristics. Finally, our results indicates potential presence of strong homophily and/or peer effects among university students.
EPJ Data Science, 2018

A common policy to affect student composition is to redraw school attendance boundaries. Yet redrawing only works if households comply and enroll in the designated school. Employing a novel dataset with unprecedented detail, we exploit changes over time in schools’ geographic attendance boundaries to provide causal estimates of how district school characteristics affect compliance with the assigned school. Households defy reassignments to schools with children from less resourceful families by enrolling in other public schools. The response to changes in school composition has a strong social gradient: resourceful households respond more to changes in school composition. We apply a boundary discontinuity design to characterize non-compliance through private school enrollment and residential relocation in the long term and once again document a strong social gradient. Our findings imply that attendance boundary policies have limited scope for desegregating schools.
-, 2018

Recent Research

. Wi-Finding: Urban Transportation Sensing Using Crowdsourced Wi-Fi. -, 2018.


. Academic Performance and Behavioral Patterns. EPJ Data Science, 2018.


. Class attendance, peer similarity, and academic performance in a large field study. PLoS One, 2017.


. Dynamics of Sorting and Inequality. -, 2016.

Recent Posts

Interview with Information about reforming high school admission, can be read here. I suggest implementing the succesful deferred …


I am a lecturer at University of Copenhagen where I teach:

  • Social Data Science. An introduction on how to apply data science and open source tools to improve research in social science. Check out the course page for materials and syllabus. See formal course information and how to register here.

  • Topics in Social Data Science. A follow up on the introductory course. We will build on the fundamentals on applying machine learning as well as tools and methods for three fundamental datatypes: networks, geo-location and text data. Course homepage here