Advanced search
1 file | 312.86 KB Add to list

Efficient and private scoring of decision trees, support vector machines and logistic regression models based on pre-computation

Author
Organization
Abstract
Many data-driven personalized services require that private data of users is scored against a trained machine learning model. In this paper we propose a novel protocol for privacy-preserving classification of decision trees, a popular machine learning model in these scenarios. Our solutions is composed out of building blocks, namely a secure comparison protocol, a protocol for obliviously selecting inputs, and a protocol for multiplication. By combining some of the building blocks for our decision tree classification protocol, we also improve previously proposed solutions for classification of support vector machines and logistic regression models. Our protocols are information theoretically secure and, unlike previously proposed solutions, do not require modular exponentiations. We show that our protocols for privacy-preserving classification lead to more efficient results from the point of view of computational and communication complexities. We present accuracy and runtime results for seven classification benchmark datasets from the UCI repository.
Keywords
Private classification, decision trees, support vector machines, logistic regression, secure multiparty computation, secret sharing, privacy-preserving computation, UNCONDITIONALLY SECURE, COMMITMENT

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 312.86 KB

Citation

Please use this url to cite or link to this publication:

MLA
De Cock, Martine et al. “Efficient and Private Scoring of Decision Trees, Support Vector Machines and Logistic Regression Models Based on Pre-computation.” IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING 16.2 (2019): 217–230. Print.
APA
De Cock, M., Dowsley, R., Horst, C., Katti, R., Nascimento, A., Poon, W.-S., & Truex, S. (2019). Efficient and private scoring of decision trees, support vector machines and logistic regression models based on pre-computation. IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING, 16(2), 217–230.
Chicago author-date
De Cock, Martine, Rafael Dowsley, Caleb Horst, Rajendra Katti, Anderson Nascimento, Wing-Sea Poon, and Stacey Truex. 2019. “Efficient and Private Scoring of Decision Trees, Support Vector Machines and Logistic Regression Models Based on Pre-computation.” Ieee Transactions on Dependable and Secure Computing 16 (2): 217–230.
Chicago author-date (all authors)
De Cock, Martine, Rafael Dowsley, Caleb Horst, Rajendra Katti, Anderson Nascimento, Wing-Sea Poon, and Stacey Truex. 2019. “Efficient and Private Scoring of Decision Trees, Support Vector Machines and Logistic Regression Models Based on Pre-computation.” Ieee Transactions on Dependable and Secure Computing 16 (2): 217–230.
Vancouver
1.
De Cock M, Dowsley R, Horst C, Katti R, Nascimento A, Poon W-S, et al. Efficient and private scoring of decision trees, support vector machines and logistic regression models based on pre-computation. IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING. 2019;16(2):217–30.
IEEE
[1]
M. De Cock et al., “Efficient and private scoring of decision trees, support vector machines and logistic regression models based on pre-computation,” IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING, vol. 16, no. 2, pp. 217–230, 2019.
@article{8607729,
  abstract     = {Many data-driven personalized services require that private data of users is scored against a trained machine learning model. In this paper we propose a novel protocol for privacy-preserving classification of decision trees, a popular machine learning model in these scenarios. Our solutions is composed out of building blocks, namely a secure comparison protocol, a protocol for obliviously selecting inputs, and a protocol for multiplication. By combining some of the building blocks for our decision tree classification protocol, we also improve previously proposed solutions for classification of support vector machines and logistic regression models. Our protocols are information theoretically secure and, unlike previously proposed solutions, do not require modular exponentiations. We show that our protocols for privacy-preserving classification lead to more efficient results from the point of view of computational and communication complexities. We present accuracy and runtime results for seven classification benchmark datasets from the UCI repository.},
  author       = {De Cock, Martine and Dowsley, Rafael and Horst, Caleb and Katti, Rajendra and Nascimento, Anderson and Poon, Wing-Sea and Truex, Stacey},
  issn         = {1545-5971},
  journal      = {IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING},
  keywords     = {Private classification,decision trees,support vector machines,logistic regression,secure multiparty computation,secret sharing,privacy-preserving computation,UNCONDITIONALLY SECURE,COMMITMENT},
  language     = {eng},
  number       = {2},
  pages        = {217--230},
  title        = {Efficient and private scoring of decision trees, support vector machines and logistic regression models based on pre-computation},
  url          = {http://dx.doi.org/10.1109/TDSC.2017.2679189},
  volume       = {16},
  year         = {2019},
}

Altmetric
View in Altmetric
Web of Science
Times cited: