Title :
An Application of Differentially Private Linear Mixed Modeling
Author :
Abowd, John M. ; Schneider, Matthew J.
Author_Institution :
Dept. of Econ., Cornell Univ., Ithaca, NY, USA
Abstract :
We consider a differentially private MLE for the linear mixed-effects model with normal random errors. This model is important because it is frequently used in small area estimation and detailed industry tabulations that present significant challenges for confidentiality protection of the underlying data. The differentially private estimator performs well compared to the regular MLE, and deteriorates as the protection increases, for a problem in which small-area variation is at the county level. More dimensions of random effects are needed to adequately represent the time-dimension of the data, and for these cases the differentially private MLE cannot be computed.
Keywords :
data mining; data privacy; statistical analysis; MLE; PPD; confidentiality protection; differentially private estimator; differentially private linear mixed modeling; linear mixed-effects model; privacy- preserving datamining; small-area variation; statistical disclosure limitation; Correlation; Data models; Industries; Maximum likelihood estimation; Privacy; Strontium; Differential Privacy; EBLUP; Linear Mixed Models; MLE; Privacy-preserving datamining; Quarterly Workforce Indicators; REML; Statistical Disclosure Limitation;
Conference_Titel :
Data Mining Workshops (ICDMW), 2011 IEEE 11th International Conference on
Conference_Location :
Vancouver, BC
Print_ISBN :
978-1-4673-0005-6
DOI :
10.1109/ICDMW.2011.26
