Record linkage and the Downstream Task
Linking data from multiple databases increases both the size and scope of a dataset, enabling post-processing tasks such as linear regression or capture-recapture to be performed. This work focused on novel approaches to multiple downstream tasks that allow for errors to propagate through the analyses and provide accurate inference. Joint work with Brenda Betancourt and Rebecca Steorts.
Computationally Scalable Bayesian Record Linkage
Linking data from multiple databases can increase the utility of many datasets and performing this linkage procedure using Bayesian methods can greatly enhance the analysis that results through the opportunity for error propagation. Unfortunately, Bayesian record linkage models are computationally complex and can be slow to fit. This work allows for greater scalability. Joint work with Neil Marchant, Rebecca Steorts, Ben Rubenstein, and Daniel N. Elazar.