Description
- Abstract:
- Probabilistic record linkage is a process to identify records in two separate datasets that represent the same entity in the absence of unique identifiers. Bayesian record linkage algorithms provide a mechanism to propagate the uncertainty of the linkage structure. An important aspect of Bayesian analysis is the specification of prior distributions for model parameters. In some applications, information on records that represent the same entities is available and can be used as prior information. However, records that are known to represent the same entities may differ from records for whom the entity is unknown. The power prior distribution is an informative prior distribution that has been proposed to incorporate historical data. We examine the performance of the power prior distribution in file linkage applications using simulation analyses. The simulations were based on data from the CDC Behavioral Risk Factor Surveillance System. Under specific simulations configurations, using the power prior distribution improved linkage accuracy. The power prior distribution appears to be a coherent and prescriptive method to incorporate known records that represent the same entities in a record linkage process.
- Notes:
- Thesis (Sc. M.)--Brown University, 2020
Citation
DeVone, Frank,
"Power Prior Distribution for Bayesian Record Linkage"
(2020).
Biostatistics Theses and Dissertations.
Brown Digital Repository. Brown University Library.
https://repository.library.brown.edu/studio/item/bdr:5eeqt566/
Relations
Collection:
-
Biostatistics Theses and Dissertations
Theses and Dissertations for the Biostatistics department....