In this unique research environment, we use machine learning (ML) to analyze Swedish population registers in order to deepen our knowledge of key social issues. With ML we refer to computationally intensive algorithms that find patterns and associations in data. ML is successfully used in diverse areas such as disease diagnostics and face recognition, but it is underutilized in the social sciences. This is unfortunate since ML has the potential to deepen our understanding of many social processes. The size, breadth, and scope of Swedish register data makes it ideally suited for ML, and ML enables us to better utilize the information contained in the registers. The research environment is located at the Institute for Analytical Sociology and consists of a renowned, international, and interdisciplinary research group. Using ML, we will develop improved neighborhood definitions; learn how individuals are differentially affected by various neighborhood characteristics at different stages of their life-course; learn more about when and for whom policy interventions have the greatest impact; and more. Combining ML and Swedish register data promises detailed and nuanced accounts of important social processes, and the combination of ML and register data places the project at the international research frontier of computational social science.
Funded by the Swedish Research Council.
Grant size: Grant in total for the project is 11.217.265 SEK. Rolf Lyneborg Lund's part is 670.000 SEK.
Start/end date: 1 January 2020 → 31 December 2023