Logistic regression with variables subject to post randomization method

Yong Ming Jeffrey Woo; Aleksandra B. Slavković

doi:10.1007/978-3-642-33627-0_10

Logistic regression with variables subject to post randomization method

Yong Ming Jeffrey Woo, Aleksandra B. Slavković

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

4 Scopus citations

Abstract

The Post Randomization Method (PRAM) is a disclosure avoidance method, where values of categorical variables are perturbed via some known probability mechanism, and only the perturbed data are released thus raising issues regarding disclosure risk and data utility. In this paper, we develop and implement a number of EM algorithms to obtain unbiased estimates of the logistic regression model with data subject to PRAM, and thus effectively account for the effects of PRAM and preserve data utility. Three different cases are considered: (1) covariates subject to PRAM, (2) response variable subject to PRAM, and (3) both covariates and response variables subject to PRAM. The proposed techniques improve on current methodology by increasing the applicability of PRAM to a wider range of products and could be extended to other type of generalized linear models. The effects of the level of perturbation and sample size on the estimates are evaluated, and relevant standard error estimates are developed and reported.

Original language	English (US)
Title of host publication	Privacy in Statistical Databases - UNESCO Chair in Data Privacy, International Conference, PSD 2012, Proceedings
Publisher	Springer Verlag
Pages	116-130
Number of pages	15
ISBN (Print)	9783642336263
DOIs	https://doi.org/10.1007/978-3-642-33627-0_10
State	Published - 2012
Event	International Conference on Privacy in Statistical Databases, PSD 2012 - Palermo, Italy Duration: Sep 26 2012 → Sep 28 2012

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	7556 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	International Conference on Privacy in Statistical Databases, PSD 2012
Country/Territory	Italy
City	Palermo
Period	9/26/12 → 9/28/12

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-642-33627-0_10

Cite this

Woo, Y. M. J., & Slavković, A. B. (2012). Logistic regression with variables subject to post randomization method. In Privacy in Statistical Databases - UNESCO Chair in Data Privacy, International Conference, PSD 2012, Proceedings (pp. 116-130). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7556 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-642-33627-0_10

Woo, Yong Ming Jeffrey ; Slavković, Aleksandra B. / Logistic regression with variables subject to post randomization method. Privacy in Statistical Databases - UNESCO Chair in Data Privacy, International Conference, PSD 2012, Proceedings. Springer Verlag, 2012. pp. 116-130 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{6106c40d7287476f827efcc767b20adf,

title = "Logistic regression with variables subject to post randomization method",

abstract = "The Post Randomization Method (PRAM) is a disclosure avoidance method, where values of categorical variables are perturbed via some known probability mechanism, and only the perturbed data are released thus raising issues regarding disclosure risk and data utility. In this paper, we develop and implement a number of EM algorithms to obtain unbiased estimates of the logistic regression model with data subject to PRAM, and thus effectively account for the effects of PRAM and preserve data utility. Three different cases are considered: (1) covariates subject to PRAM, (2) response variable subject to PRAM, and (3) both covariates and response variables subject to PRAM. The proposed techniques improve on current methodology by increasing the applicability of PRAM to a wider range of products and could be extended to other type of generalized linear models. The effects of the level of perturbation and sample size on the estimates are evaluated, and relevant standard error estimates are developed and reported.",

author = "Woo, {Yong Ming Jeffrey} and Slavkovi{\'c}, {Aleksandra B.}",

year = "2012",

doi = "10.1007/978-3-642-33627-0_10",

language = "English (US)",

isbn = "9783642336263",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "116--130",

booktitle = "Privacy in Statistical Databases - UNESCO Chair in Data Privacy, International Conference, PSD 2012, Proceedings",

address = "Germany",

}

Woo, YMJ & Slavković, AB 2012, Logistic regression with variables subject to post randomization method. in Privacy in Statistical Databases - UNESCO Chair in Data Privacy, International Conference, PSD 2012, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7556 LNCS, Springer Verlag, pp. 116-130, International Conference on Privacy in Statistical Databases, PSD 2012, Palermo, Italy, 9/26/12. https://doi.org/10.1007/978-3-642-33627-0_10

Logistic regression with variables subject to post randomization method. / Woo, Yong Ming Jeffrey; Slavković, Aleksandra B.
Privacy in Statistical Databases - UNESCO Chair in Data Privacy, International Conference, PSD 2012, Proceedings. Springer Verlag, 2012. p. 116-130 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7556 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Logistic regression with variables subject to post randomization method

AU - Woo, Yong Ming Jeffrey

AU - Slavković, Aleksandra B.

PY - 2012

Y1 - 2012

N2 - The Post Randomization Method (PRAM) is a disclosure avoidance method, where values of categorical variables are perturbed via some known probability mechanism, and only the perturbed data are released thus raising issues regarding disclosure risk and data utility. In this paper, we develop and implement a number of EM algorithms to obtain unbiased estimates of the logistic regression model with data subject to PRAM, and thus effectively account for the effects of PRAM and preserve data utility. Three different cases are considered: (1) covariates subject to PRAM, (2) response variable subject to PRAM, and (3) both covariates and response variables subject to PRAM. The proposed techniques improve on current methodology by increasing the applicability of PRAM to a wider range of products and could be extended to other type of generalized linear models. The effects of the level of perturbation and sample size on the estimates are evaluated, and relevant standard error estimates are developed and reported.

AB - The Post Randomization Method (PRAM) is a disclosure avoidance method, where values of categorical variables are perturbed via some known probability mechanism, and only the perturbed data are released thus raising issues regarding disclosure risk and data utility. In this paper, we develop and implement a number of EM algorithms to obtain unbiased estimates of the logistic regression model with data subject to PRAM, and thus effectively account for the effects of PRAM and preserve data utility. Three different cases are considered: (1) covariates subject to PRAM, (2) response variable subject to PRAM, and (3) both covariates and response variables subject to PRAM. The proposed techniques improve on current methodology by increasing the applicability of PRAM to a wider range of products and could be extended to other type of generalized linear models. The effects of the level of perturbation and sample size on the estimates are evaluated, and relevant standard error estimates are developed and reported.

UR - http://www.scopus.com/inward/record.url?scp=84867509153&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84867509153&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-33627-0_10

DO - 10.1007/978-3-642-33627-0_10

M3 - Conference contribution

AN - SCOPUS:84867509153

SN - 9783642336263

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 116

EP - 130

BT - Privacy in Statistical Databases - UNESCO Chair in Data Privacy, International Conference, PSD 2012, Proceedings

PB - Springer Verlag

T2 - International Conference on Privacy in Statistical Databases, PSD 2012

Y2 - 26 September 2012 through 28 September 2012

ER -

Woo YMJ, Slavković AB. Logistic regression with variables subject to post randomization method. In Privacy in Statistical Databases - UNESCO Chair in Data Privacy, International Conference, PSD 2012, Proceedings. Springer Verlag. 2012. p. 116-130. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-33627-0_10

Logistic regression with variables subject to post randomization method

Abstract

Publication series

Other

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this